base.c 86.6 KB
Newer Older
1
// SPDX-License-Identifier: GPL-2.0
L
Linus Torvalds 已提交
2 3 4 5 6 7 8 9 10 11 12 13 14
/*
 *  linux/fs/proc/base.c
 *
 *  Copyright (C) 1991, 1992 Linus Torvalds
 *
 *  proc base directory handling functions
 *
 *  1999, Al Viro. Rewritten. Now it covers the whole per-process part.
 *  Instead of using magical inumbers to determine the kind of object
 *  we allocate and fill in-core inodes upon lookup. They don't even
 *  go into icache. We cache the reference to task_struct upon lookup too.
 *  Eventually it should become a filesystem in its own. We don't use the
 *  rest of procfs anymore.
M
Mauricio Lin 已提交
15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48
 *
 *
 *  Changelog:
 *  17-Jan-2005
 *  Allan Bezerra
 *  Bruna Moreira <bruna.moreira@indt.org.br>
 *  Edjard Mota <edjard.mota@indt.org.br>
 *  Ilias Biris <ilias.biris@indt.org.br>
 *  Mauricio Lin <mauricio.lin@indt.org.br>
 *
 *  Embedded Linux Lab - 10LE Instituto Nokia de Tecnologia - INdT
 *
 *  A new process specific entry (smaps) included in /proc. It shows the
 *  size of rss for each memory area. The maps entry lacks information
 *  about physical memory size (rss) for each mapped file, i.e.,
 *  rss information for executables and library files.
 *  This additional information is useful for any tools that need to know
 *  about physical memory consumption for a process specific library.
 *
 *  Changelog:
 *  21-Feb-2005
 *  Embedded Linux Lab - 10LE Instituto Nokia de Tecnologia - INdT
 *  Pud inclusion in the page table walking.
 *
 *  ChangeLog:
 *  10-Mar-2005
 *  10LE Instituto Nokia de Tecnologia - INdT:
 *  A better way to walks through the page table as suggested by Hugh Dickins.
 *
 *  Simo Piiroinen <simo.piiroinen@nokia.com>:
 *  Smaps information related to shared, private, clean and dirty pages.
 *
 *  Paul Mundt <paul.mundt@nokia.com>:
 *  Overall revision about smaps.
L
Linus Torvalds 已提交
49 50
 */

51
#include <linux/uaccess.h>
L
Linus Torvalds 已提交
52 53 54 55 56

#include <linux/errno.h>
#include <linux/time.h>
#include <linux/proc_fs.h>
#include <linux/stat.h>
57
#include <linux/task_io_accounting_ops.h>
L
Linus Torvalds 已提交
58
#include <linux/init.h>
59
#include <linux/capability.h>
L
Linus Torvalds 已提交
60
#include <linux/file.h>
A
Al Viro 已提交
61
#include <linux/fdtable.h>
L
Linus Torvalds 已提交
62 63 64
#include <linux/string.h>
#include <linux/seq_file.h>
#include <linux/namei.h>
65
#include <linux/mnt_namespace.h>
L
Linus Torvalds 已提交
66
#include <linux/mm.h>
D
David Rientjes 已提交
67
#include <linux/swap.h>
68
#include <linux/rcupdate.h>
L
Linus Torvalds 已提交
69
#include <linux/kallsyms.h>
K
Ken Chen 已提交
70
#include <linux/stacktrace.h>
71
#include <linux/resource.h>
K
Kees Cook 已提交
72
#include <linux/module.h>
L
Linus Torvalds 已提交
73 74 75
#include <linux/mount.h>
#include <linux/security.h>
#include <linux/ptrace.h>
76
#include <linux/tracehook.h>
A
Andrew Morton 已提交
77
#include <linux/printk.h>
A
Alexey Dobriyan 已提交
78
#include <linux/cache.h>
79
#include <linux/cgroup.h>
L
Linus Torvalds 已提交
80 81
#include <linux/cpuset.h>
#include <linux/audit.h>
A
Al Viro 已提交
82
#include <linux/poll.h>
83
#include <linux/nsproxy.h>
84
#include <linux/oom.h>
85
#include <linux/elf.h>
86
#include <linux/pid_namespace.h>
87
#include <linux/user_namespace.h>
88
#include <linux/fs_struct.h>
89
#include <linux/slab.h>
90
#include <linux/sched/autogroup.h>
91
#include <linux/sched/mm.h>
92
#include <linux/sched/coredump.h>
93
#include <linux/sched/debug.h>
94
#include <linux/sched/stat.h>
95
#include <linux/flex_array.h>
96
#include <linux/posix-timers.h>
97
#include <trace/events/oom.h>
L
Linus Torvalds 已提交
98
#include "internal.h"
99
#include "fd.h"
L
Linus Torvalds 已提交
100

101 102
#include "../../lib/kstrtox.h"

103 104 105 106 107 108 109 110 111 112
/* NOTE:
 *	Implementing inode permission operations in /proc is almost
 *	certainly an error.  Permission checks need to happen during
 *	each system call not at open time.  The reason is that most of
 *	what we wish to check for permissions in /proc varies at runtime.
 *
 *	The classic example of a problem is opening file descriptors
 *	in /proc for a task before it execs a suid executable.
 */

A
Alexey Dobriyan 已提交
113 114
static u8 nlink_tid __ro_after_init;
static u8 nlink_tgid __ro_after_init;
115

L
Linus Torvalds 已提交
116
struct pid_entry {
117
	const char *name;
118
	unsigned int len;
A
Al Viro 已提交
119
	umode_t mode;
120
	const struct inode_operations *iop;
121
	const struct file_operations *fop;
122
	union proc_op op;
L
Linus Torvalds 已提交
123 124
};

125
#define NOD(NAME, MODE, IOP, FOP, OP) {			\
126
	.name = (NAME),					\
127
	.len  = sizeof(NAME) - 1,			\
128 129 130 131 132 133
	.mode = MODE,					\
	.iop  = IOP,					\
	.fop  = FOP,					\
	.op   = OP,					\
}

A
Alexey Dobriyan 已提交
134 135 136
#define DIR(NAME, MODE, iops, fops)	\
	NOD(NAME, (S_IFDIR|(MODE)), &iops, &fops, {} )
#define LNK(NAME, get_link)					\
137
	NOD(NAME, (S_IFLNK|S_IRWXUGO),				\
138
		&proc_pid_link_inode_operations, NULL,		\
A
Alexey Dobriyan 已提交
139 140 141 142
		{ .proc_get_link = get_link } )
#define REG(NAME, MODE, fops)				\
	NOD(NAME, (S_IFREG|(MODE)), NULL, &fops, {})
#define ONE(NAME, MODE, show)				\
143 144
	NOD(NAME, (S_IFREG|(MODE)), 			\
		NULL, &proc_single_file_operations,	\
A
Alexey Dobriyan 已提交
145
		{ .proc_show = show } )
L
Linus Torvalds 已提交
146

147 148 149 150
/*
 * Count the number of hardlinks for the pid_entry table, excluding the .
 * and .. links.
 */
151
static unsigned int __init pid_entry_nlink(const struct pid_entry *entries,
152 153 154 155 156
	unsigned int n)
{
	unsigned int i;
	unsigned int count;

157
	count = 2;
158 159 160 161 162 163 164 165
	for (i = 0; i < n; ++i) {
		if (S_ISDIR(entries[i].mode))
			++count;
	}

	return count;
}

166
static int get_task_root(struct task_struct *task, struct path *root)
L
Linus Torvalds 已提交
167
{
H
Hugh Dickins 已提交
168 169
	int result = -ENOENT;

170
	task_lock(task);
171 172
	if (task->fs) {
		get_fs_root(task->fs, root);
H
Hugh Dickins 已提交
173 174
		result = 0;
	}
175
	task_unlock(task);
H
Hugh Dickins 已提交
176
	return result;
177 178
}

179
static int proc_cwd_link(struct dentry *dentry, struct path *path)
180
{
181
	struct task_struct *task = get_proc_task(d_inode(dentry));
182
	int result = -ENOENT;
183 184

	if (task) {
185 186 187 188 189 190
		task_lock(task);
		if (task->fs) {
			get_fs_pwd(task->fs, path);
			result = 0;
		}
		task_unlock(task);
191 192
		put_task_struct(task);
	}
L
Linus Torvalds 已提交
193 194 195
	return result;
}

196
static int proc_root_link(struct dentry *dentry, struct path *path)
L
Linus Torvalds 已提交
197
{
198
	struct task_struct *task = get_proc_task(d_inode(dentry));
L
Linus Torvalds 已提交
199
	int result = -ENOENT;
200 201

	if (task) {
202
		result = get_task_root(task, path);
203 204
		put_task_struct(task);
	}
L
Linus Torvalds 已提交
205 206 207
	return result;
}

208
static ssize_t get_mm_cmdline(struct mm_struct *mm, char __user *buf,
209
			      size_t count, loff_t *ppos)
L
Linus Torvalds 已提交
210
{
211
	unsigned long arg_start, arg_end, env_start, env_end;
212 213
	unsigned long pos, len;
	char *page;
214 215

	/* Check if process spawned far enough to have cmdline. */
216 217
	if (!mm->env_end)
		return 0;
218

219
	spin_lock(&mm->arg_lock);
220 221 222 223
	arg_start = mm->arg_start;
	arg_end = mm->arg_end;
	env_start = mm->env_start;
	env_end = mm->env_end;
224
	spin_unlock(&mm->arg_lock);
225

226 227
	if (arg_start >= arg_end)
		return 0;
228

229
	/*
230 231 232 233
	 * We have traditionally allowed the user to re-write
	 * the argument strings and overflow the end result
	 * into the environment section. But only do that if
	 * the environment area is contiguous to the arguments.
234
	 */
235 236 237
	if (env_start != arg_end || env_start >= env_end)
		env_start = env_end = arg_end;

238 239 240 241
	/* .. and limit it to a maximum of one page of slop */
	if (env_end >= arg_end + PAGE_SIZE)
		env_end = arg_end + PAGE_SIZE - 1;

242 243
	/* We're not going to care if "*ppos" has high bits set */
	pos = arg_start + *ppos;
244

245 246 247
	/* .. but we do check the result is in the proper range */
	if (pos < arg_start || pos >= env_end)
		return 0;
248

249 250 251 252 253 254 255 256 257 258 259 260
	/* .. and we never go past env_end */
	if (env_end - pos < count)
		count = env_end - pos;

	page = (char *)__get_free_page(GFP_KERNEL);
	if (!page)
		return -ENOMEM;

	len = 0;
	while (count) {
		int got;
		size_t size = min_t(size_t, PAGE_SIZE, count);
261 262 263 264 265 266 267 268
		long offset;

		/*
		 * Are we already starting past the official end?
		 * We always include the last byte that is *supposed*
		 * to be NUL
		 */
		offset = (pos >= arg_end) ? pos - arg_end + 1 : 0;
269

270 271
		got = access_remote_vm(mm, pos - offset, page, size + offset, FOLL_ANON);
		if (got <= offset)
272
			break;
273
		got -= offset;
274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291

		/* Don't walk past a NUL character once you hit arg_end */
		if (pos + got >= arg_end) {
			int n = 0;

			/*
			 * If we started before 'arg_end' but ended up
			 * at or after it, we start the NUL character
			 * check at arg_end-1 (where we expect the normal
			 * EOF to be).
			 *
			 * NOTE! This is smaller than 'got', because
			 * pos + got >= arg_end
			 */
			if (pos < arg_end)
				n = arg_end - pos - 1;

			/* Cut off at first NUL after 'n' */
292 293
			got = n + strnlen(page+n, offset+got-n);
			if (got < offset)
294
				break;
295 296 297 298 299
			got -= offset;

			/* Include the NUL if it existed */
			if (got < size)
				got++;
300
		}
301

302
		got -= copy_to_user(buf, page+offset, got);
303 304 305 306
		if (unlikely(!got)) {
			if (!len)
				len = -EFAULT;
			break;
307
		}
308 309 310 311
		pos += got;
		buf += got;
		len += got;
		count -= got;
312 313 314
	}

	free_page((unsigned long)page);
315
	return len;
L
Linus Torvalds 已提交
316 317
}

318 319 320 321 322 323 324 325 326 327 328
static ssize_t get_task_cmdline(struct task_struct *tsk, char __user *buf,
				size_t count, loff_t *pos)
{
	struct mm_struct *mm;
	ssize_t ret;

	mm = get_task_mm(tsk);
	if (!mm)
		return 0;

	ret = get_mm_cmdline(mm, buf, count, pos);
329
	mmput(mm);
330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348
	return ret;
}

static ssize_t proc_pid_cmdline_read(struct file *file, char __user *buf,
				     size_t count, loff_t *pos)
{
	struct task_struct *tsk;
	ssize_t ret;

	BUG_ON(*pos < 0);

	tsk = get_proc_task(file_inode(file));
	if (!tsk)
		return -ESRCH;
	ret = get_task_cmdline(tsk, buf, count, pos);
	put_task_struct(tsk);
	if (ret > 0)
		*pos += ret;
	return ret;
L
Linus Torvalds 已提交
349 350
}

351 352 353 354 355
static const struct file_operations proc_pid_cmdline_ops = {
	.read	= proc_pid_cmdline_read,
	.llseek	= generic_file_llseek,
};

L
Linus Torvalds 已提交
356 357 358 359 360
#ifdef CONFIG_KALLSYMS
/*
 * Provides a wchan file via kallsyms in a proper one-value-per-file format.
 * Returns the resolved symbol.  If that fails, simply return the address.
 */
361 362
static int proc_pid_wchan(struct seq_file *m, struct pid_namespace *ns,
			  struct pid *pid, struct task_struct *task)
L
Linus Torvalds 已提交
363
{
A
Alexey Dobriyan 已提交
364
	unsigned long wchan;
365
	char symname[KSYM_NAME_LEN];
L
Linus Torvalds 已提交
366

367 368
	if (!ptrace_may_access(task, PTRACE_MODE_READ_FSCREDS))
		goto print0;
L
Linus Torvalds 已提交
369

370 371
	wchan = get_wchan(task);
	if (wchan && !lookup_symbol_name(wchan, symname)) {
372
		seq_puts(m, symname);
373 374
		return 0;
	}
375

376 377
print0:
	seq_putc(m, '0');
378
	return 0;
L
Linus Torvalds 已提交
379 380 381
}
#endif /* CONFIG_KALLSYMS */

382 383 384 385 386
static int lock_trace(struct task_struct *task)
{
	int err = mutex_lock_killable(&task->signal->cred_guard_mutex);
	if (err)
		return err;
387
	if (!ptrace_may_access(task, PTRACE_MODE_ATTACH_FSCREDS)) {
388 389 390 391 392 393 394 395 396 397 398
		mutex_unlock(&task->signal->cred_guard_mutex);
		return -EPERM;
	}
	return 0;
}

static void unlock_trace(struct task_struct *task)
{
	mutex_unlock(&task->signal->cred_guard_mutex);
}

K
Ken Chen 已提交
399 400 401 402 403 404 405 406 407
#ifdef CONFIG_STACKTRACE

#define MAX_STACK_TRACE_DEPTH	64

static int proc_pid_stack(struct seq_file *m, struct pid_namespace *ns,
			  struct pid *pid, struct task_struct *task)
{
	struct stack_trace trace;
	unsigned long *entries;
408
	int err;
K
Ken Chen 已提交
409

410 411 412 413 414 415 416 417 418 419 420 421 422 423
	/*
	 * The ability to racily run the kernel stack unwinder on a running task
	 * and then observe the unwinder output is scary; while it is useful for
	 * debugging kernel issues, it can also allow an attacker to leak kernel
	 * stack contents.
	 * Doing this in a manner that is at least safe from races would require
	 * some work to ensure that the remote task can not be scheduled; and
	 * even then, this would still expose the unwinder as local attack
	 * surface.
	 * Therefore, this interface is restricted to root.
	 */
	if (!file_ns_capable(m->file, &init_user_ns, CAP_SYS_ADMIN))
		return -EACCES;

424 425
	entries = kmalloc_array(MAX_STACK_TRACE_DEPTH, sizeof(*entries),
				GFP_KERNEL);
K
Ken Chen 已提交
426 427 428 429 430 431 432 433
	if (!entries)
		return -ENOMEM;

	trace.nr_entries	= 0;
	trace.max_entries	= MAX_STACK_TRACE_DEPTH;
	trace.entries		= entries;
	trace.skip		= 0;

434 435
	err = lock_trace(task);
	if (!err) {
436 437
		unsigned int i;

438 439 440
		save_stack_trace_tsk(task, &trace);

		for (i = 0; i < trace.nr_entries; i++) {
441
			seq_printf(m, "[<0>] %pB\n", (void *)entries[i]);
442 443
		}
		unlock_trace(task);
K
Ken Chen 已提交
444 445 446
	}
	kfree(entries);

447
	return err;
K
Ken Chen 已提交
448 449 450
}
#endif

451
#ifdef CONFIG_SCHED_INFO
L
Linus Torvalds 已提交
452 453 454
/*
 * Provides /proc/PID/schedstat
 */
455 456
static int proc_pid_schedstat(struct seq_file *m, struct pid_namespace *ns,
			      struct pid *pid, struct task_struct *task)
L
Linus Torvalds 已提交
457
{
458 459 460 461
	if (unlikely(!sched_info_on()))
		seq_printf(m, "0 0 0\n");
	else
		seq_printf(m, "%llu %llu %lu\n",
462 463 464 465 466
		   (unsigned long long)task->se.sum_exec_runtime,
		   (unsigned long long)task->sched_info.run_delay,
		   task->sched_info.pcount);

	return 0;
L
Linus Torvalds 已提交
467 468 469
}
#endif

A
Arjan van de Ven 已提交
470 471 472 473
#ifdef CONFIG_LATENCYTOP
static int lstats_show_proc(struct seq_file *m, void *v)
{
	int i;
474 475
	struct inode *inode = m->private;
	struct task_struct *task = get_proc_task(inode);
A
Arjan van de Ven 已提交
476

477 478 479
	if (!task)
		return -ESRCH;
	seq_puts(m, "Latency Top version : v0.1\n");
480
	for (i = 0; i < LT_SAVECOUNT; i++) {
481 482
		struct latency_record *lr = &task->latency_record[i];
		if (lr->backtrace[0]) {
A
Arjan van de Ven 已提交
483
			int q;
484 485
			seq_printf(m, "%i %li %li",
				   lr->count, lr->time, lr->max);
A
Arjan van de Ven 已提交
486
			for (q = 0; q < LT_BACKTRACEDEPTH; q++) {
487 488
				unsigned long bt = lr->backtrace[q];
				if (!bt)
A
Arjan van de Ven 已提交
489
					break;
490
				if (bt == ULONG_MAX)
A
Arjan van de Ven 已提交
491
					break;
492
				seq_printf(m, " %ps", (void *)bt);
A
Arjan van de Ven 已提交
493
			}
494
			seq_putc(m, '\n');
A
Arjan van de Ven 已提交
495 496 497
		}

	}
498
	put_task_struct(task);
A
Arjan van de Ven 已提交
499 500 501 502 503
	return 0;
}

static int lstats_open(struct inode *inode, struct file *file)
{
504
	return single_open(file, lstats_show_proc, inode);
505 506
}

A
Arjan van de Ven 已提交
507 508 509
static ssize_t lstats_write(struct file *file, const char __user *buf,
			    size_t count, loff_t *offs)
{
A
Al Viro 已提交
510
	struct task_struct *task = get_proc_task(file_inode(file));
A
Arjan van de Ven 已提交
511

512 513
	if (!task)
		return -ESRCH;
A
Arjan van de Ven 已提交
514
	clear_all_latency_tracing(task);
515
	put_task_struct(task);
A
Arjan van de Ven 已提交
516 517 518 519 520 521 522 523 524

	return count;
}

static const struct file_operations proc_lstats_operations = {
	.open		= lstats_open,
	.read		= seq_read,
	.write		= lstats_write,
	.llseek		= seq_lseek,
525
	.release	= single_release,
A
Arjan van de Ven 已提交
526 527 528 529
};

#endif

530 531
static int proc_oom_score(struct seq_file *m, struct pid_namespace *ns,
			  struct pid *pid, struct task_struct *task)
L
Linus Torvalds 已提交
532
{
533
	unsigned long totalpages = totalram_pages() + total_swap_pages;
534
	unsigned long points = 0;
L
Linus Torvalds 已提交
535

536 537
	points = oom_badness(task, NULL, NULL, totalpages) *
					1000 / totalpages;
538 539 540
	seq_printf(m, "%lu\n", points);

	return 0;
L
Linus Torvalds 已提交
541 542
}

543
struct limit_names {
544 545
	const char *name;
	const char *unit;
546 547 548
};

static const struct limit_names lnames[RLIM_NLIMITS] = {
549
	[RLIMIT_CPU] = {"Max cpu time", "seconds"},
550 551 552 553 554 555 556 557 558 559 560 561 562 563
	[RLIMIT_FSIZE] = {"Max file size", "bytes"},
	[RLIMIT_DATA] = {"Max data size", "bytes"},
	[RLIMIT_STACK] = {"Max stack size", "bytes"},
	[RLIMIT_CORE] = {"Max core file size", "bytes"},
	[RLIMIT_RSS] = {"Max resident set", "bytes"},
	[RLIMIT_NPROC] = {"Max processes", "processes"},
	[RLIMIT_NOFILE] = {"Max open files", "files"},
	[RLIMIT_MEMLOCK] = {"Max locked memory", "bytes"},
	[RLIMIT_AS] = {"Max address space", "bytes"},
	[RLIMIT_LOCKS] = {"Max file locks", "locks"},
	[RLIMIT_SIGPENDING] = {"Max pending signals", "signals"},
	[RLIMIT_MSGQUEUE] = {"Max msgqueue size", "bytes"},
	[RLIMIT_NICE] = {"Max nice priority", NULL},
	[RLIMIT_RTPRIO] = {"Max realtime priority", NULL},
564
	[RLIMIT_RTTIME] = {"Max realtime timeout", "us"},
565 566 567
};

/* Display limits for a process */
568 569
static int proc_pid_limits(struct seq_file *m, struct pid_namespace *ns,
			   struct pid *pid, struct task_struct *task)
570 571 572 573 574 575
{
	unsigned int i;
	unsigned long flags;

	struct rlimit rlim[RLIM_NLIMITS];

576
	if (!lock_task_sighand(task, &flags))
577 578 579 580 581 582 583
		return 0;
	memcpy(rlim, task->signal->rlim, sizeof(struct rlimit) * RLIM_NLIMITS);
	unlock_task_sighand(task, &flags);

	/*
	 * print the file header
	 */
584 585 586 587
	seq_puts(m, "Limit                     "
		"Soft Limit           "
		"Hard Limit           "
		"Units     \n");
588 589 590

	for (i = 0; i < RLIM_NLIMITS; i++) {
		if (rlim[i].rlim_cur == RLIM_INFINITY)
591
			seq_printf(m, "%-25s %-20s ",
592
				   lnames[i].name, "unlimited");
593
		else
594
			seq_printf(m, "%-25s %-20lu ",
595
				   lnames[i].name, rlim[i].rlim_cur);
596 597

		if (rlim[i].rlim_max == RLIM_INFINITY)
598
			seq_printf(m, "%-20s ", "unlimited");
599
		else
600
			seq_printf(m, "%-20lu ", rlim[i].rlim_max);
601 602

		if (lnames[i].unit)
603
			seq_printf(m, "%-10s\n", lnames[i].unit);
604
		else
605
			seq_putc(m, '\n');
606 607
	}

608
	return 0;
609 610
}

R
Roland McGrath 已提交
611
#ifdef CONFIG_HAVE_ARCH_TRACEHOOK
612 613
static int proc_pid_syscall(struct seq_file *m, struct pid_namespace *ns,
			    struct pid *pid, struct task_struct *task)
R
Roland McGrath 已提交
614 615 616
{
	long nr;
	unsigned long args[6], sp, pc;
617 618 619
	int res;

	res = lock_trace(task);
620 621
	if (res)
		return res;
R
Roland McGrath 已提交
622 623

	if (task_current_syscall(task, &nr, args, 6, &sp, &pc))
624
		seq_puts(m, "running\n");
625
	else if (nr < 0)
626
		seq_printf(m, "%ld 0x%lx 0x%lx\n", nr, sp, pc);
627
	else
628
		seq_printf(m,
R
Roland McGrath 已提交
629 630 631 632
		       "%ld 0x%lx 0x%lx 0x%lx 0x%lx 0x%lx 0x%lx 0x%lx 0x%lx\n",
		       nr,
		       args[0], args[1], args[2], args[3], args[4], args[5],
		       sp, pc);
633
	unlock_trace(task);
634 635

	return 0;
R
Roland McGrath 已提交
636 637 638
}
#endif /* CONFIG_HAVE_ARCH_TRACEHOOK */

L
Linus Torvalds 已提交
639 640 641 642 643
/************************************************************************/
/*                       Here the fs part begins                        */
/************************************************************************/

/* permission checks */
644
static int proc_fd_access_allowed(struct inode *inode)
L
Linus Torvalds 已提交
645
{
646 647
	struct task_struct *task;
	int allowed = 0;
648 649 650
	/* Allow access to a task's file descriptors if it is us or we
	 * may use ptrace attach to the process and find out that
	 * information.
651 652
	 */
	task = get_proc_task(inode);
653
	if (task) {
654
		allowed = ptrace_may_access(task, PTRACE_MODE_READ_FSCREDS);
655
		put_task_struct(task);
656
	}
657
	return allowed;
L
Linus Torvalds 已提交
658 659
}

660
int proc_setattr(struct dentry *dentry, struct iattr *attr)
661 662
{
	int error;
663
	struct inode *inode = d_inode(dentry);
664 665 666 667

	if (attr->ia_valid & ATTR_MODE)
		return -EPERM;

668
	error = setattr_prepare(dentry, attr);
C
Christoph Hellwig 已提交
669 670 671 672 673 674
	if (error)
		return error;

	setattr_copy(inode, attr);
	mark_inode_dirty(inode);
	return 0;
675 676
}

677 678 679 680 681 682 683 684 685 686 687 688
/*
 * May current process learn task's sched/cmdline info (for hide_pid_min=1)
 * or euid/egid (for hide_pid_min=2)?
 */
static bool has_pid_permissions(struct pid_namespace *pid,
				 struct task_struct *task,
				 int hide_pid_min)
{
	if (pid->hide_pid < hide_pid_min)
		return true;
	if (in_group_p(pid->pid_gid))
		return true;
689
	return ptrace_may_access(task, PTRACE_MODE_READ_FSCREDS);
690 691 692 693 694
}


static int proc_pid_permission(struct inode *inode, int mask)
{
695
	struct pid_namespace *pid = proc_pid_ns(inode);
696 697 698 699
	struct task_struct *task;
	bool has_perms;

	task = get_proc_task(inode);
700 701
	if (!task)
		return -ESRCH;
702
	has_perms = has_pid_permissions(pid, task, HIDEPID_NO_ACCESS);
703 704 705
	put_task_struct(task);

	if (!has_perms) {
706
		if (pid->hide_pid == HIDEPID_INVISIBLE) {
707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722
			/*
			 * Let's make getdents(), stat(), and open()
			 * consistent with each other.  If a process
			 * may not stat() a file, it shouldn't be seen
			 * in procfs at all.
			 */
			return -ENOENT;
		}

		return -EPERM;
	}
	return generic_permission(inode, mask);
}



723
static const struct inode_operations proc_def_inode_operations = {
724 725 726
	.setattr	= proc_setattr,
};

727 728 729
static int proc_single_show(struct seq_file *m, void *v)
{
	struct inode *inode = m->private;
730 731
	struct pid_namespace *ns = proc_pid_ns(inode);
	struct pid *pid = proc_pid(inode);
732 733 734 735 736 737 738 739 740 741 742 743 744 745 746
	struct task_struct *task;
	int ret;

	task = get_pid_task(pid, PIDTYPE_PID);
	if (!task)
		return -ESRCH;

	ret = PROC_I(inode)->op.proc_show(m, ns, pid, task);

	put_task_struct(task);
	return ret;
}

static int proc_single_open(struct inode *inode, struct file *filp)
{
J
Jovi Zhang 已提交
747
	return single_open(filp, proc_single_show, inode);
748 749 750 751 752 753 754 755 756
}

static const struct file_operations proc_single_file_operations = {
	.open		= proc_single_open,
	.read		= seq_read,
	.llseek		= seq_lseek,
	.release	= single_release,
};

O
Oleg Nesterov 已提交
757 758

struct mm_struct *proc_mem_open(struct inode *inode, unsigned int mode)
L
Linus Torvalds 已提交
759
{
O
Oleg Nesterov 已提交
760 761
	struct task_struct *task = get_proc_task(inode);
	struct mm_struct *mm = ERR_PTR(-ESRCH);
762

O
Oleg Nesterov 已提交
763
	if (task) {
764
		mm = mm_access(task, mode | PTRACE_MODE_FSCREDS);
O
Oleg Nesterov 已提交
765
		put_task_struct(task);
766

O
Oleg Nesterov 已提交
767 768
		if (!IS_ERR_OR_NULL(mm)) {
			/* ensure this mm_struct can't be freed */
V
Vegard Nossum 已提交
769
			mmgrab(mm);
O
Oleg Nesterov 已提交
770 771 772 773 774 775 776 777 778 779 780
			/* but do not pin its memory */
			mmput(mm);
		}
	}

	return mm;
}

static int __mem_open(struct inode *inode, struct file *file, unsigned int mode)
{
	struct mm_struct *mm = proc_mem_open(inode, mode);
781 782 783 784 785

	if (IS_ERR(mm))
		return PTR_ERR(mm);

	file->private_data = mm;
L
Linus Torvalds 已提交
786 787 788
	return 0;
}

789 790
static int mem_open(struct inode *inode, struct file *file)
{
791 792 793 794 795 796
	int ret = __mem_open(inode, file, PTRACE_MODE_ATTACH);

	/* OK to pass negative loff_t, we can catch out-of-range */
	file->f_mode |= FMODE_UNSIGNED_OFFSET;

	return ret;
797 798
}

799 800
static ssize_t mem_rw(struct file *file, char __user *buf,
			size_t count, loff_t *ppos, int write)
L
Linus Torvalds 已提交
801
{
802
	struct mm_struct *mm = file->private_data;
803 804
	unsigned long addr = *ppos;
	ssize_t copied;
L
Linus Torvalds 已提交
805
	char *page;
806
	unsigned int flags;
L
Linus Torvalds 已提交
807

808 809
	if (!mm)
		return 0;
810

811
	page = (char *)__get_free_page(GFP_KERNEL);
812
	if (!page)
813
		return -ENOMEM;
L
Linus Torvalds 已提交
814

815
	copied = 0;
V
Vegard Nossum 已提交
816
	if (!mmget_not_zero(mm))
817 818
		goto free;

L
Linus Torvalds 已提交
819
	flags = FOLL_FORCE | (write ? FOLL_WRITE : 0);
820

L
Linus Torvalds 已提交
821
	while (count > 0) {
822
		int this_len = min_t(int, count, PAGE_SIZE);
L
Linus Torvalds 已提交
823

824
		if (write && copy_from_user(page, buf, this_len)) {
L
Linus Torvalds 已提交
825 826 827
			copied = -EFAULT;
			break;
		}
828

829
		this_len = access_remote_vm(mm, addr, page, this_len, flags);
830
		if (!this_len) {
L
Linus Torvalds 已提交
831 832 833 834
			if (!copied)
				copied = -EIO;
			break;
		}
835 836 837 838 839 840 841 842 843 844

		if (!write && copy_to_user(buf, page, this_len)) {
			copied = -EFAULT;
			break;
		}

		buf += this_len;
		addr += this_len;
		copied += this_len;
		count -= this_len;
L
Linus Torvalds 已提交
845
	}
846
	*ppos = addr;
847

848 849
	mmput(mm);
free:
850
	free_page((unsigned long) page);
L
Linus Torvalds 已提交
851 852 853
	return copied;
}

854 855 856 857 858 859 860 861 862 863 864 865
static ssize_t mem_read(struct file *file, char __user *buf,
			size_t count, loff_t *ppos)
{
	return mem_rw(file, buf, count, ppos, 0);
}

static ssize_t mem_write(struct file *file, const char __user *buf,
			 size_t count, loff_t *ppos)
{
	return mem_rw(file, (char __user*)buf, count, ppos, 1);
}

866
loff_t mem_lseek(struct file *file, loff_t offset, int orig)
L
Linus Torvalds 已提交
867 868 869 870 871 872 873 874 875 876 877 878 879 880 881
{
	switch (orig) {
	case 0:
		file->f_pos = offset;
		break;
	case 1:
		file->f_pos += offset;
		break;
	default:
		return -EINVAL;
	}
	force_successful_syscall_return();
	return file->f_pos;
}

882 883 884
static int mem_release(struct inode *inode, struct file *file)
{
	struct mm_struct *mm = file->private_data;
885
	if (mm)
886
		mmdrop(mm);
887 888 889
	return 0;
}

890
static const struct file_operations proc_mem_operations = {
L
Linus Torvalds 已提交
891 892 893 894
	.llseek		= mem_lseek,
	.read		= mem_read,
	.write		= mem_write,
	.open		= mem_open,
895
	.release	= mem_release,
L
Linus Torvalds 已提交
896 897
};

898 899 900 901 902
static int environ_open(struct inode *inode, struct file *file)
{
	return __mem_open(inode, file, PTRACE_MODE_READ);
}

903 904 905 906 907
static ssize_t environ_read(struct file *file, char __user *buf,
			size_t count, loff_t *ppos)
{
	char *page;
	unsigned long src = *ppos;
908 909
	int ret = 0;
	struct mm_struct *mm = file->private_data;
910
	unsigned long env_start, env_end;
911

912 913
	/* Ensure the process spawned far enough to have an environment. */
	if (!mm || !mm->env_end)
914
		return 0;
915

916
	page = (char *)__get_free_page(GFP_KERNEL);
917
	if (!page)
918
		return -ENOMEM;
919

A
Al Viro 已提交
920
	ret = 0;
V
Vegard Nossum 已提交
921
	if (!mmget_not_zero(mm))
922
		goto free;
923

924
	spin_lock(&mm->arg_lock);
925 926
	env_start = mm->env_start;
	env_end = mm->env_end;
927
	spin_unlock(&mm->arg_lock);
928

929
	while (count > 0) {
930 931
		size_t this_len, max_len;
		int retval;
932

933
		if (src >= (env_end - env_start))
934 935
			break;

936
		this_len = env_end - (env_start + src);
937 938 939

		max_len = min_t(size_t, PAGE_SIZE, count);
		this_len = min(max_len, this_len);
940

941
		retval = access_remote_vm(mm, (env_start + src), page, this_len, FOLL_ANON);
942 943 944 945 946 947 948 949 950 951 952 953 954 955 956 957 958 959

		if (retval <= 0) {
			ret = retval;
			break;
		}

		if (copy_to_user(buf, page, retval)) {
			ret = -EFAULT;
			break;
		}

		ret += retval;
		src += retval;
		buf += retval;
		count -= retval;
	}
	*ppos = src;
	mmput(mm);
960 961

free:
962 963 964 965 966
	free_page((unsigned long) page);
	return ret;
}

static const struct file_operations proc_environ_operations = {
967
	.open		= environ_open,
968
	.read		= environ_read,
969
	.llseek		= generic_file_llseek,
970
	.release	= mem_release,
971 972
};

A
Al Viro 已提交
973 974 975 976 977 978 979 980 981 982
static int auxv_open(struct inode *inode, struct file *file)
{
	return __mem_open(inode, file, PTRACE_MODE_READ_FSCREDS);
}

static ssize_t auxv_read(struct file *file, char __user *buf,
			size_t count, loff_t *ppos)
{
	struct mm_struct *mm = file->private_data;
	unsigned int nwords = 0;
983 984 985

	if (!mm)
		return 0;
A
Al Viro 已提交
986 987 988 989 990 991 992 993 994 995 996 997 998 999
	do {
		nwords += 2;
	} while (mm->saved_auxv[nwords - 2] != 0); /* AT_NULL */
	return simple_read_from_buffer(buf, count, ppos, mm->saved_auxv,
				       nwords * sizeof(mm->saved_auxv[0]));
}

static const struct file_operations proc_auxv_operations = {
	.open		= auxv_open,
	.read		= auxv_read,
	.llseek		= generic_file_llseek,
	.release	= mem_release,
};

1000 1001 1002
static ssize_t oom_adj_read(struct file *file, char __user *buf, size_t count,
			    loff_t *ppos)
{
A
Al Viro 已提交
1003
	struct task_struct *task = get_proc_task(file_inode(file));
1004 1005 1006 1007 1008 1009
	char buffer[PROC_NUMBUF];
	int oom_adj = OOM_ADJUST_MIN;
	size_t len;

	if (!task)
		return -ESRCH;
1010 1011 1012 1013 1014
	if (task->signal->oom_score_adj == OOM_SCORE_ADJ_MAX)
		oom_adj = OOM_ADJUST_MAX;
	else
		oom_adj = (task->signal->oom_score_adj * -OOM_DISABLE) /
			  OOM_SCORE_ADJ_MAX;
1015 1016 1017 1018 1019
	put_task_struct(task);
	len = snprintf(buffer, sizeof(buffer), "%d\n", oom_adj);
	return simple_read_from_buffer(buf, count, ppos, buffer, len);
}

1020 1021 1022
static int __set_oom_adj(struct file *file, int oom_adj, bool legacy)
{
	static DEFINE_MUTEX(oom_adj_mutex);
1023
	struct mm_struct *mm = NULL;
1024 1025 1026 1027 1028 1029 1030 1031 1032 1033 1034 1035 1036 1037 1038 1039 1040 1041 1042 1043 1044 1045 1046 1047 1048 1049 1050 1051 1052
	struct task_struct *task;
	int err = 0;

	task = get_proc_task(file_inode(file));
	if (!task)
		return -ESRCH;

	mutex_lock(&oom_adj_mutex);
	if (legacy) {
		if (oom_adj < task->signal->oom_score_adj &&
				!capable(CAP_SYS_RESOURCE)) {
			err = -EACCES;
			goto err_unlock;
		}
		/*
		 * /proc/pid/oom_adj is provided for legacy purposes, ask users to use
		 * /proc/pid/oom_score_adj instead.
		 */
		pr_warn_once("%s (%d): /proc/%d/oom_adj is deprecated, please use /proc/%d/oom_score_adj instead.\n",
			  current->comm, task_pid_nr(current), task_pid_nr(task),
			  task_pid_nr(task));
	} else {
		if ((short)oom_adj < task->signal->oom_score_adj_min &&
				!capable(CAP_SYS_RESOURCE)) {
			err = -EACCES;
			goto err_unlock;
		}
	}

1053 1054 1055 1056 1057 1058 1059 1060 1061 1062 1063
	/*
	 * Make sure we will check other processes sharing the mm if this is
	 * not vfrok which wants its own oom_score_adj.
	 * pin the mm so it doesn't go away and get reused after task_unlock
	 */
	if (!task->vfork_done) {
		struct task_struct *p = find_lock_task_mm(task);

		if (p) {
			if (atomic_read(&p->mm->mm_users) > 1) {
				mm = p->mm;
V
Vegard Nossum 已提交
1064
				mmgrab(mm);
1065 1066 1067 1068 1069
			}
			task_unlock(p);
		}
	}

1070 1071 1072 1073
	task->signal->oom_score_adj = oom_adj;
	if (!legacy && has_capability_noaudit(current, CAP_SYS_RESOURCE))
		task->signal->oom_score_adj_min = (short)oom_adj;
	trace_oom_score_adj_update(task);
1074 1075 1076 1077 1078 1079 1080 1081 1082 1083 1084 1085 1086 1087 1088 1089 1090 1091 1092 1093 1094 1095 1096 1097

	if (mm) {
		struct task_struct *p;

		rcu_read_lock();
		for_each_process(p) {
			if (same_thread_group(task, p))
				continue;

			/* do not touch kernel threads or the global init */
			if (p->flags & PF_KTHREAD || is_global_init(p))
				continue;

			task_lock(p);
			if (!p->vfork_done && process_shares_mm(p, mm)) {
				p->signal->oom_score_adj = oom_adj;
				if (!legacy && has_capability_noaudit(current, CAP_SYS_RESOURCE))
					p->signal->oom_score_adj_min = (short)oom_adj;
			}
			task_unlock(p);
		}
		rcu_read_unlock();
		mmdrop(mm);
	}
1098 1099 1100 1101 1102
err_unlock:
	mutex_unlock(&oom_adj_mutex);
	put_task_struct(task);
	return err;
}
1103

1104 1105 1106 1107 1108 1109 1110 1111 1112 1113
/*
 * /proc/pid/oom_adj exists solely for backwards compatibility with previous
 * kernels.  The effective policy is defined by oom_score_adj, which has a
 * different scale: oom_adj grew exponentially and oom_score_adj grows linearly.
 * Values written to oom_adj are simply mapped linearly to oom_score_adj.
 * Processes that become oom disabled via oom_adj will still be oom disabled
 * with this implementation.
 *
 * oom_adj cannot be removed since existing userspace binaries use it.
 */
1114 1115 1116 1117 1118 1119 1120 1121 1122 1123 1124 1125 1126 1127 1128 1129 1130 1131 1132 1133 1134 1135 1136 1137 1138 1139 1140 1141 1142 1143 1144 1145 1146
static ssize_t oom_adj_write(struct file *file, const char __user *buf,
			     size_t count, loff_t *ppos)
{
	char buffer[PROC_NUMBUF];
	int oom_adj;
	int err;

	memset(buffer, 0, sizeof(buffer));
	if (count > sizeof(buffer) - 1)
		count = sizeof(buffer) - 1;
	if (copy_from_user(buffer, buf, count)) {
		err = -EFAULT;
		goto out;
	}

	err = kstrtoint(strstrip(buffer), 0, &oom_adj);
	if (err)
		goto out;
	if ((oom_adj < OOM_ADJUST_MIN || oom_adj > OOM_ADJUST_MAX) &&
	     oom_adj != OOM_DISABLE) {
		err = -EINVAL;
		goto out;
	}

	/*
	 * Scale /proc/pid/oom_score_adj appropriately ensuring that a maximum
	 * value is always attainable.
	 */
	if (oom_adj == OOM_ADJUST_MAX)
		oom_adj = OOM_SCORE_ADJ_MAX;
	else
		oom_adj = (oom_adj * OOM_SCORE_ADJ_MAX) / -OOM_DISABLE;

1147
	err = __set_oom_adj(file, oom_adj, true);
1148 1149 1150 1151 1152 1153 1154 1155 1156 1157
out:
	return err < 0 ? err : count;
}

static const struct file_operations proc_oom_adj_operations = {
	.read		= oom_adj_read,
	.write		= oom_adj_write,
	.llseek		= generic_file_llseek,
};

D
David Rientjes 已提交
1158 1159 1160
static ssize_t oom_score_adj_read(struct file *file, char __user *buf,
					size_t count, loff_t *ppos)
{
A
Al Viro 已提交
1161
	struct task_struct *task = get_proc_task(file_inode(file));
D
David Rientjes 已提交
1162
	char buffer[PROC_NUMBUF];
1163
	short oom_score_adj = OOM_SCORE_ADJ_MIN;
D
David Rientjes 已提交
1164 1165 1166 1167
	size_t len;

	if (!task)
		return -ESRCH;
1168
	oom_score_adj = task->signal->oom_score_adj;
D
David Rientjes 已提交
1169
	put_task_struct(task);
1170
	len = snprintf(buffer, sizeof(buffer), "%hd\n", oom_score_adj);
D
David Rientjes 已提交
1171 1172 1173 1174 1175 1176 1177
	return simple_read_from_buffer(buf, count, ppos, buffer, len);
}

static ssize_t oom_score_adj_write(struct file *file, const char __user *buf,
					size_t count, loff_t *ppos)
{
	char buffer[PROC_NUMBUF];
A
Alexey Dobriyan 已提交
1178
	int oom_score_adj;
D
David Rientjes 已提交
1179 1180 1181 1182 1183
	int err;

	memset(buffer, 0, sizeof(buffer));
	if (count > sizeof(buffer) - 1)
		count = sizeof(buffer) - 1;
1184 1185 1186 1187
	if (copy_from_user(buffer, buf, count)) {
		err = -EFAULT;
		goto out;
	}
D
David Rientjes 已提交
1188

A
Alexey Dobriyan 已提交
1189
	err = kstrtoint(strstrip(buffer), 0, &oom_score_adj);
D
David Rientjes 已提交
1190
	if (err)
1191
		goto out;
D
David Rientjes 已提交
1192
	if (oom_score_adj < OOM_SCORE_ADJ_MIN ||
1193 1194 1195 1196
			oom_score_adj > OOM_SCORE_ADJ_MAX) {
		err = -EINVAL;
		goto out;
	}
D
David Rientjes 已提交
1197

1198
	err = __set_oom_adj(file, oom_score_adj, false);
1199 1200
out:
	return err < 0 ? err : count;
D
David Rientjes 已提交
1201 1202 1203 1204 1205
}

static const struct file_operations proc_oom_score_adj_operations = {
	.read		= oom_score_adj_read,
	.write		= oom_score_adj_write,
1206
	.llseek		= default_llseek,
D
David Rientjes 已提交
1207 1208
};

L
Linus Torvalds 已提交
1209
#ifdef CONFIG_AUDITSYSCALL
1210
#define TMPBUFLEN 11
L
Linus Torvalds 已提交
1211 1212 1213
static ssize_t proc_loginuid_read(struct file * file, char __user * buf,
				  size_t count, loff_t *ppos)
{
A
Al Viro 已提交
1214
	struct inode * inode = file_inode(file);
1215
	struct task_struct *task = get_proc_task(inode);
L
Linus Torvalds 已提交
1216 1217 1218
	ssize_t length;
	char tmpbuf[TMPBUFLEN];

1219 1220
	if (!task)
		return -ESRCH;
L
Linus Torvalds 已提交
1221
	length = scnprintf(tmpbuf, TMPBUFLEN, "%u",
1222 1223
			   from_kuid(file->f_cred->user_ns,
				     audit_get_loginuid(task)));
1224
	put_task_struct(task);
L
Linus Torvalds 已提交
1225 1226 1227 1228 1229 1230
	return simple_read_from_buffer(buf, count, ppos, tmpbuf, length);
}

static ssize_t proc_loginuid_write(struct file * file, const char __user * buf,
				   size_t count, loff_t *ppos)
{
A
Al Viro 已提交
1231
	struct inode * inode = file_inode(file);
L
Linus Torvalds 已提交
1232
	uid_t loginuid;
1233
	kuid_t kloginuid;
1234
	int rv;
L
Linus Torvalds 已提交
1235

1236 1237 1238
	rcu_read_lock();
	if (current != pid_task(proc_pid(inode), PIDTYPE_PID)) {
		rcu_read_unlock();
L
Linus Torvalds 已提交
1239
		return -EPERM;
1240 1241
	}
	rcu_read_unlock();
L
Linus Torvalds 已提交
1242 1243 1244 1245 1246 1247

	if (*ppos != 0) {
		/* No partial writes. */
		return -EINVAL;
	}

1248 1249 1250
	rv = kstrtou32_from_user(buf, count, 10, &loginuid);
	if (rv < 0)
		return rv;
1251 1252 1253 1254 1255 1256

	/* is userspace tring to explicitly UNSET the loginuid? */
	if (loginuid == AUDIT_UID_UNSET) {
		kloginuid = INVALID_UID;
	} else {
		kloginuid = make_kuid(file->f_cred->user_ns, loginuid);
1257 1258
		if (!uid_valid(kloginuid))
			return -EINVAL;
1259 1260
	}

1261 1262 1263 1264
	rv = audit_set_loginuid(kloginuid);
	if (rv < 0)
		return rv;
	return count;
L
Linus Torvalds 已提交
1265 1266
}

1267
static const struct file_operations proc_loginuid_operations = {
L
Linus Torvalds 已提交
1268 1269
	.read		= proc_loginuid_read,
	.write		= proc_loginuid_write,
1270
	.llseek		= generic_file_llseek,
L
Linus Torvalds 已提交
1271
};
1272 1273 1274 1275

static ssize_t proc_sessionid_read(struct file * file, char __user * buf,
				  size_t count, loff_t *ppos)
{
A
Al Viro 已提交
1276
	struct inode * inode = file_inode(file);
1277 1278 1279 1280 1281 1282 1283 1284 1285 1286 1287 1288 1289 1290
	struct task_struct *task = get_proc_task(inode);
	ssize_t length;
	char tmpbuf[TMPBUFLEN];

	if (!task)
		return -ESRCH;
	length = scnprintf(tmpbuf, TMPBUFLEN, "%u",
				audit_get_sessionid(task));
	put_task_struct(task);
	return simple_read_from_buffer(buf, count, ppos, tmpbuf, length);
}

static const struct file_operations proc_sessionid_operations = {
	.read		= proc_sessionid_read,
1291
	.llseek		= generic_file_llseek,
1292
};
L
Linus Torvalds 已提交
1293 1294
#endif

1295 1296 1297 1298
#ifdef CONFIG_FAULT_INJECTION
static ssize_t proc_fault_inject_read(struct file * file, char __user * buf,
				      size_t count, loff_t *ppos)
{
A
Al Viro 已提交
1299
	struct task_struct *task = get_proc_task(file_inode(file));
1300 1301 1302 1303 1304 1305 1306 1307 1308 1309
	char buffer[PROC_NUMBUF];
	size_t len;
	int make_it_fail;

	if (!task)
		return -ESRCH;
	make_it_fail = task->make_it_fail;
	put_task_struct(task);

	len = snprintf(buffer, sizeof(buffer), "%i\n", make_it_fail);
1310 1311

	return simple_read_from_buffer(buf, count, ppos, buffer, len);
1312 1313 1314 1315 1316 1317
}

static ssize_t proc_fault_inject_write(struct file * file,
			const char __user * buf, size_t count, loff_t *ppos)
{
	struct task_struct *task;
1318
	char buffer[PROC_NUMBUF];
1319
	int make_it_fail;
1320
	int rv;
1321 1322 1323 1324 1325 1326 1327 1328

	if (!capable(CAP_SYS_RESOURCE))
		return -EPERM;
	memset(buffer, 0, sizeof(buffer));
	if (count > sizeof(buffer) - 1)
		count = sizeof(buffer) - 1;
	if (copy_from_user(buffer, buf, count))
		return -EFAULT;
1329 1330 1331
	rv = kstrtoint(strstrip(buffer), 0, &make_it_fail);
	if (rv < 0)
		return rv;
1332 1333 1334
	if (make_it_fail < 0 || make_it_fail > 1)
		return -EINVAL;

A
Al Viro 已提交
1335
	task = get_proc_task(file_inode(file));
1336 1337 1338 1339
	if (!task)
		return -ESRCH;
	task->make_it_fail = make_it_fail;
	put_task_struct(task);
1340 1341

	return count;
1342 1343
}

1344
static const struct file_operations proc_fault_inject_operations = {
1345 1346
	.read		= proc_fault_inject_read,
	.write		= proc_fault_inject_write,
1347
	.llseek		= generic_file_llseek,
1348
};
1349 1350 1351 1352 1353

static ssize_t proc_fail_nth_write(struct file *file, const char __user *buf,
				   size_t count, loff_t *ppos)
{
	struct task_struct *task;
1354 1355
	int err;
	unsigned int n;
1356

1357 1358 1359 1360
	err = kstrtouint_from_user(buf, count, 0, &n);
	if (err)
		return err;

1361 1362 1363
	task = get_proc_task(file_inode(file));
	if (!task)
		return -ESRCH;
1364
	task->fail_nth = n;
1365
	put_task_struct(task);
1366

1367 1368 1369 1370 1371 1372 1373
	return count;
}

static ssize_t proc_fail_nth_read(struct file *file, char __user *buf,
				  size_t count, loff_t *ppos)
{
	struct task_struct *task;
1374 1375
	char numbuf[PROC_NUMBUF];
	ssize_t len;
1376 1377 1378 1379

	task = get_proc_task(file_inode(file));
	if (!task)
		return -ESRCH;
1380
	len = snprintf(numbuf, sizeof(numbuf), "%u\n", task->fail_nth);
1381
	put_task_struct(task);
1382
	return simple_read_from_buffer(buf, count, ppos, numbuf, len);
1383 1384 1385 1386 1387 1388
}

static const struct file_operations proc_fail_nth_operations = {
	.read		= proc_fail_nth_read,
	.write		= proc_fail_nth_write,
};
1389 1390
#endif

A
Arjan van de Ven 已提交
1391

I
Ingo Molnar 已提交
1392 1393 1394 1395 1396 1397 1398
#ifdef CONFIG_SCHED_DEBUG
/*
 * Print out various scheduling related per-task fields:
 */
static int sched_show(struct seq_file *m, void *v)
{
	struct inode *inode = m->private;
1399
	struct pid_namespace *ns = proc_pid_ns(inode);
I
Ingo Molnar 已提交
1400 1401 1402 1403 1404
	struct task_struct *p;

	p = get_proc_task(inode);
	if (!p)
		return -ESRCH;
1405
	proc_sched_show_task(p, ns, m);
I
Ingo Molnar 已提交
1406 1407 1408 1409 1410 1411 1412 1413 1414 1415

	put_task_struct(p);

	return 0;
}

static ssize_t
sched_write(struct file *file, const char __user *buf,
	    size_t count, loff_t *offset)
{
A
Al Viro 已提交
1416
	struct inode *inode = file_inode(file);
I
Ingo Molnar 已提交
1417 1418 1419 1420 1421 1422 1423 1424 1425 1426 1427 1428 1429 1430
	struct task_struct *p;

	p = get_proc_task(inode);
	if (!p)
		return -ESRCH;
	proc_sched_set_task(p);

	put_task_struct(p);

	return count;
}

static int sched_open(struct inode *inode, struct file *filp)
{
J
Jovi Zhang 已提交
1431
	return single_open(filp, sched_show, inode);
I
Ingo Molnar 已提交
1432 1433 1434 1435 1436 1437 1438
}

static const struct file_operations proc_pid_sched_operations = {
	.open		= sched_open,
	.read		= seq_read,
	.write		= sched_write,
	.llseek		= seq_lseek,
1439
	.release	= single_release,
I
Ingo Molnar 已提交
1440 1441 1442 1443
};

#endif

1444 1445 1446 1447 1448 1449 1450 1451 1452 1453 1454 1455 1456 1457 1458 1459 1460 1461 1462 1463 1464 1465 1466
#ifdef CONFIG_SCHED_AUTOGROUP
/*
 * Print out autogroup related information:
 */
static int sched_autogroup_show(struct seq_file *m, void *v)
{
	struct inode *inode = m->private;
	struct task_struct *p;

	p = get_proc_task(inode);
	if (!p)
		return -ESRCH;
	proc_sched_autogroup_show_task(p, m);

	put_task_struct(p);

	return 0;
}

static ssize_t
sched_autogroup_write(struct file *file, const char __user *buf,
	    size_t count, loff_t *offset)
{
A
Al Viro 已提交
1467
	struct inode *inode = file_inode(file);
1468 1469
	struct task_struct *p;
	char buffer[PROC_NUMBUF];
A
Alexey Dobriyan 已提交
1470
	int nice;
1471 1472 1473 1474 1475 1476 1477 1478
	int err;

	memset(buffer, 0, sizeof(buffer));
	if (count > sizeof(buffer) - 1)
		count = sizeof(buffer) - 1;
	if (copy_from_user(buffer, buf, count))
		return -EFAULT;

A
Alexey Dobriyan 已提交
1479 1480 1481
	err = kstrtoint(strstrip(buffer), 0, &nice);
	if (err < 0)
		return err;
1482 1483 1484 1485 1486

	p = get_proc_task(inode);
	if (!p)
		return -ESRCH;

1487
	err = proc_sched_autogroup_set_nice(p, nice);
1488 1489 1490 1491 1492 1493 1494 1495 1496 1497 1498 1499 1500 1501 1502 1503 1504 1505 1506 1507 1508 1509 1510 1511 1512 1513 1514 1515 1516 1517 1518
	if (err)
		count = err;

	put_task_struct(p);

	return count;
}

static int sched_autogroup_open(struct inode *inode, struct file *filp)
{
	int ret;

	ret = single_open(filp, sched_autogroup_show, NULL);
	if (!ret) {
		struct seq_file *m = filp->private_data;

		m->private = inode;
	}
	return ret;
}

static const struct file_operations proc_pid_sched_autogroup_operations = {
	.open		= sched_autogroup_open,
	.read		= seq_read,
	.write		= sched_autogroup_write,
	.llseek		= seq_lseek,
	.release	= single_release,
};

#endif /* CONFIG_SCHED_AUTOGROUP */

1519 1520 1521
static ssize_t comm_write(struct file *file, const char __user *buf,
				size_t count, loff_t *offset)
{
A
Al Viro 已提交
1522
	struct inode *inode = file_inode(file);
1523 1524
	struct task_struct *p;
	char buffer[TASK_COMM_LEN];
1525
	const size_t maxlen = sizeof(buffer) - 1;
1526 1527

	memset(buffer, 0, sizeof(buffer));
1528
	if (copy_from_user(buffer, buf, count > maxlen ? maxlen : count))
1529 1530 1531 1532 1533 1534 1535 1536 1537 1538 1539 1540 1541 1542 1543 1544 1545 1546 1547 1548 1549 1550 1551 1552 1553
		return -EFAULT;

	p = get_proc_task(inode);
	if (!p)
		return -ESRCH;

	if (same_thread_group(current, p))
		set_task_comm(p, buffer);
	else
		count = -EINVAL;

	put_task_struct(p);

	return count;
}

static int comm_show(struct seq_file *m, void *v)
{
	struct inode *inode = m->private;
	struct task_struct *p;

	p = get_proc_task(inode);
	if (!p)
		return -ESRCH;

1554 1555
	proc_task_name(m, p, false);
	seq_putc(m, '\n');
1556 1557 1558 1559 1560 1561 1562 1563

	put_task_struct(p);

	return 0;
}

static int comm_open(struct inode *inode, struct file *filp)
{
J
Jovi Zhang 已提交
1564
	return single_open(filp, comm_show, inode);
1565 1566 1567 1568 1569 1570 1571 1572 1573 1574
}

static const struct file_operations proc_pid_set_comm_operations = {
	.open		= comm_open,
	.read		= seq_read,
	.write		= comm_write,
	.llseek		= seq_lseek,
	.release	= single_release,
};

1575
static int proc_exe_link(struct dentry *dentry, struct path *exe_path)
M
Matt Helsley 已提交
1576 1577 1578 1579
{
	struct task_struct *task;
	struct file *exe_file;

1580
	task = get_proc_task(d_inode(dentry));
M
Matt Helsley 已提交
1581 1582
	if (!task)
		return -ENOENT;
M
Mateusz Guzik 已提交
1583
	exe_file = get_task_exe_file(task);
M
Matt Helsley 已提交
1584 1585 1586 1587 1588 1589 1590 1591 1592 1593
	put_task_struct(task);
	if (exe_file) {
		*exe_path = exe_file->f_path;
		path_get(&exe_file->f_path);
		fput(exe_file);
		return 0;
	} else
		return -ENOENT;
}

1594
static const char *proc_pid_get_link(struct dentry *dentry,
1595 1596
				     struct inode *inode,
				     struct delayed_call *done)
L
Linus Torvalds 已提交
1597
{
1598
	struct path path;
L
Linus Torvalds 已提交
1599 1600
	int error = -EACCES;

1601 1602 1603
	if (!dentry)
		return ERR_PTR(-ECHILD);

1604 1605
	/* Are we allowed to snoop on the tasks file descriptors? */
	if (!proc_fd_access_allowed(inode))
L
Linus Torvalds 已提交
1606 1607
		goto out;

1608 1609 1610 1611
	error = PROC_I(inode)->op.proc_get_link(dentry, &path);
	if (error)
		goto out;

1612
	nd_jump_link(&path);
1613
	return NULL;
L
Linus Torvalds 已提交
1614
out:
1615
	return ERR_PTR(error);
L
Linus Torvalds 已提交
1616 1617
}

1618
static int do_proc_readlink(struct path *path, char __user *buffer, int buflen)
L
Linus Torvalds 已提交
1619
{
1620
	char *tmp = (char *)__get_free_page(GFP_KERNEL);
1621
	char *pathname;
L
Linus Torvalds 已提交
1622 1623 1624 1625
	int len;

	if (!tmp)
		return -ENOMEM;
1626

1627
	pathname = d_path(path, tmp, PAGE_SIZE);
1628 1629
	len = PTR_ERR(pathname);
	if (IS_ERR(pathname))
L
Linus Torvalds 已提交
1630
		goto out;
1631
	len = tmp + PAGE_SIZE - 1 - pathname;
L
Linus Torvalds 已提交
1632 1633 1634

	if (len > buflen)
		len = buflen;
1635
	if (copy_to_user(buffer, pathname, len))
L
Linus Torvalds 已提交
1636 1637 1638 1639 1640 1641 1642 1643 1644
		len = -EFAULT;
 out:
	free_page((unsigned long)tmp);
	return len;
}

static int proc_pid_readlink(struct dentry * dentry, char __user * buffer, int buflen)
{
	int error = -EACCES;
1645
	struct inode *inode = d_inode(dentry);
1646
	struct path path;
L
Linus Torvalds 已提交
1647

1648 1649
	/* Are we allowed to snoop on the tasks file descriptors? */
	if (!proc_fd_access_allowed(inode))
L
Linus Torvalds 已提交
1650 1651
		goto out;

1652
	error = PROC_I(inode)->op.proc_get_link(dentry, &path);
L
Linus Torvalds 已提交
1653 1654 1655
	if (error)
		goto out;

1656 1657
	error = do_proc_readlink(&path, buffer, buflen);
	path_put(&path);
L
Linus Torvalds 已提交
1658 1659 1660 1661
out:
	return error;
}

1662
const struct inode_operations proc_pid_link_inode_operations = {
L
Linus Torvalds 已提交
1663
	.readlink	= proc_pid_readlink,
1664
	.get_link	= proc_pid_get_link,
1665
	.setattr	= proc_setattr,
L
Linus Torvalds 已提交
1666 1667
};

1668 1669 1670

/* building an inode */

A
Al Viro 已提交
1671
void task_dump_owner(struct task_struct *task, umode_t mode,
1672 1673 1674 1675 1676 1677 1678 1679 1680
		     kuid_t *ruid, kgid_t *rgid)
{
	/* Depending on the state of dumpable compute who should own a
	 * proc file for a task.
	 */
	const struct cred *cred;
	kuid_t uid;
	kgid_t gid;

1681 1682 1683 1684 1685 1686
	if (unlikely(task->flags & PF_KTHREAD)) {
		*ruid = GLOBAL_ROOT_UID;
		*rgid = GLOBAL_ROOT_GID;
		return;
	}

1687 1688 1689 1690 1691 1692 1693 1694 1695 1696 1697 1698 1699 1700 1701 1702 1703 1704 1705 1706 1707 1708 1709 1710 1711 1712 1713 1714 1715 1716 1717 1718 1719 1720 1721 1722 1723 1724 1725 1726 1727 1728
	/* Default to the tasks effective ownership */
	rcu_read_lock();
	cred = __task_cred(task);
	uid = cred->euid;
	gid = cred->egid;
	rcu_read_unlock();

	/*
	 * Before the /proc/pid/status file was created the only way to read
	 * the effective uid of a /process was to stat /proc/pid.  Reading
	 * /proc/pid/status is slow enough that procps and other packages
	 * kept stating /proc/pid.  To keep the rules in /proc simple I have
	 * made this apply to all per process world readable and executable
	 * directories.
	 */
	if (mode != (S_IFDIR|S_IRUGO|S_IXUGO)) {
		struct mm_struct *mm;
		task_lock(task);
		mm = task->mm;
		/* Make non-dumpable tasks owned by some root */
		if (mm) {
			if (get_dumpable(mm) != SUID_DUMP_USER) {
				struct user_namespace *user_ns = mm->user_ns;

				uid = make_kuid(user_ns, 0);
				if (!uid_valid(uid))
					uid = GLOBAL_ROOT_UID;

				gid = make_kgid(user_ns, 0);
				if (!gid_valid(gid))
					gid = GLOBAL_ROOT_GID;
			}
		} else {
			uid = GLOBAL_ROOT_UID;
			gid = GLOBAL_ROOT_GID;
		}
		task_unlock(task);
	}
	*ruid = uid;
	*rgid = gid;
}

1729 1730
struct inode *proc_pid_make_inode(struct super_block * sb,
				  struct task_struct *task, umode_t mode)
1731 1732 1733
{
	struct inode * inode;
	struct proc_inode *ei;
L
Linus Torvalds 已提交
1734

1735
	/* We need a new inode */
L
Linus Torvalds 已提交
1736

1737 1738 1739 1740 1741 1742
	inode = new_inode(sb);
	if (!inode)
		goto out;

	/* Common stuff */
	ei = PROC_I(inode);
1743
	inode->i_mode = mode;
1744
	inode->i_ino = get_next_ino();
1745
	inode->i_mtime = inode->i_atime = inode->i_ctime = current_time(inode);
1746 1747 1748 1749 1750
	inode->i_op = &proc_def_inode_operations;

	/*
	 * grab the reference to task.
	 */
1751
	ei->pid = get_task_pid(task, PIDTYPE_PID);
1752 1753 1754
	if (!ei->pid)
		goto out_unlock;

1755
	task_dump_owner(task, 0, &inode->i_uid, &inode->i_gid);
1756 1757
	security_task_to_inode(task, inode);

L
Linus Torvalds 已提交
1758
out:
1759 1760 1761 1762 1763
	return inode;

out_unlock:
	iput(inode);
	return NULL;
L
Linus Torvalds 已提交
1764 1765
}

1766 1767
int pid_getattr(const struct path *path, struct kstat *stat,
		u32 request_mask, unsigned int query_flags)
L
Linus Torvalds 已提交
1768
{
1769
	struct inode *inode = d_inode(path->dentry);
1770
	struct pid_namespace *pid = proc_pid_ns(inode);
1771
	struct task_struct *task;
1772

1773
	generic_fillattr(inode, stat);
L
Linus Torvalds 已提交
1774

1775 1776
	stat->uid = GLOBAL_ROOT_UID;
	stat->gid = GLOBAL_ROOT_GID;
1777
	rcu_read_lock();
1778 1779
	task = pid_task(proc_pid(inode), PIDTYPE_PID);
	if (task) {
1780
		if (!has_pid_permissions(pid, task, HIDEPID_INVISIBLE)) {
1781 1782 1783 1784 1785 1786 1787
			rcu_read_unlock();
			/*
			 * This doesn't prevent learning whether PID exists,
			 * it only makes getattr() consistent with readdir().
			 */
			return -ENOENT;
		}
1788
		task_dump_owner(task, inode->i_mode, &stat->uid, &stat->gid);
L
Linus Torvalds 已提交
1789
	}
1790
	rcu_read_unlock();
A
Alan Cox 已提交
1791
	return 0;
L
Linus Torvalds 已提交
1792 1793 1794 1795 1796
}

/* dentry stuff */

/*
1797 1798 1799 1800 1801 1802 1803 1804 1805 1806 1807
 * Set <pid>/... inode ownership (can change due to setuid(), etc.)
 */
void pid_update_inode(struct task_struct *task, struct inode *inode)
{
	task_dump_owner(task, inode->i_mode, &inode->i_uid, &inode->i_gid);

	inode->i_mode &= ~(S_ISUID | S_ISGID);
	security_task_to_inode(task, inode);
}

/*
L
Linus Torvalds 已提交
1808 1809
 * Rewrite the inode's ownerships here because the owning task may have
 * performed a setuid(), etc.
1810
 *
L
Linus Torvalds 已提交
1811
 */
1812
static int pid_revalidate(struct dentry *dentry, unsigned int flags)
L
Linus Torvalds 已提交
1813
{
1814 1815
	struct inode *inode;
	struct task_struct *task;
1816

1817
	if (flags & LOOKUP_RCU)
1818 1819
		return -ECHILD;

1820
	inode = d_inode(dentry);
1821 1822
	task = get_proc_task(inode);

1823
	if (task) {
1824
		pid_update_inode(task, inode);
1825
		put_task_struct(task);
L
Linus Torvalds 已提交
1826 1827 1828 1829 1830
		return 1;
	}
	return 0;
}

1831 1832 1833 1834 1835
static inline bool proc_inode_is_dead(struct inode *inode)
{
	return !proc_pid(inode)->tasks[PIDTYPE_PID].first;
}

1836 1837 1838 1839 1840 1841
int pid_delete_dentry(const struct dentry *dentry)
{
	/* Is the task we represent dead?
	 * If so, then don't put the dentry on the lru list,
	 * kill it immediately.
	 */
1842
	return proc_inode_is_dead(d_inode(dentry));
1843 1844
}

1845
const struct dentry_operations pid_dentry_operations =
1846 1847 1848 1849 1850 1851 1852
{
	.d_revalidate	= pid_revalidate,
	.d_delete	= pid_delete_dentry,
};

/* Lookups */

1853 1854 1855 1856 1857 1858 1859 1860 1861 1862 1863 1864
/*
 * Fill a directory entry.
 *
 * If possible create the dcache entry and derive our inode number and
 * file type from dcache entry.
 *
 * Since all of the proc inode numbers are dynamically generated, the inode
 * numbers do not exist until the inode is cache.  This means creating the
 * the dcache entry in readdir is necessary to keep the inode numbers
 * reported by readdir in sync with the inode numbers reported
 * by stat.
 */
A
Al Viro 已提交
1865
bool proc_fill_cache(struct file *file, struct dir_context *ctx,
1866
	const char *name, unsigned int len,
1867
	instantiate_t instantiate, struct task_struct *task, const void *ptr)
1868
{
A
Al Viro 已提交
1869
	struct dentry *child, *dir = file->f_path.dentry;
1870
	struct qstr qname = QSTR_INIT(name, len);
1871
	struct inode *inode;
1872 1873
	unsigned type = DT_UNKNOWN;
	ino_t ino = 1;
1874

1875
	child = d_hash_and_lookup(dir, &qname);
1876
	if (!child) {
1877 1878 1879
		DECLARE_WAIT_QUEUE_HEAD_ONSTACK(wq);
		child = d_alloc_parallel(dir, &qname, &wq);
		if (IS_ERR(child))
1880
			goto end_instantiate;
1881
		if (d_in_lookup(child)) {
1882 1883
			struct dentry *res;
			res = instantiate(child, task, ptr);
1884
			d_lookup_done(child);
1885 1886 1887
			if (unlikely(res)) {
				dput(child);
				child = res;
1888 1889
				if (IS_ERR(child))
					goto end_instantiate;
1890
			}
1891 1892
		}
	}
1893
	inode = d_inode(child);
1894 1895
	ino = inode->i_ino;
	type = inode->i_mode >> 12;
1896
	dput(child);
1897
end_instantiate:
A
Al Viro 已提交
1898
	return dir_emit(ctx, name, len, ino, type);
1899 1900
}

1901 1902 1903 1904 1905 1906 1907
/*
 * dname_to_vma_addr - maps a dentry name into two unsigned longs
 * which represent vma start and end addresses.
 */
static int dname_to_vma_addr(struct dentry *dentry,
			     unsigned long *start, unsigned long *end)
{
1908 1909 1910 1911
	const char *str = dentry->d_name.name;
	unsigned long long sval, eval;
	unsigned int len;

1912 1913
	if (str[0] == '0' && str[1] != '-')
		return -EINVAL;
1914 1915 1916 1917 1918 1919 1920 1921
	len = _parse_integer(str, 16, &sval);
	if (len & KSTRTOX_OVERFLOW)
		return -EINVAL;
	if (sval != (unsigned long)sval)
		return -EINVAL;
	str += len;

	if (*str != '-')
1922
		return -EINVAL;
1923 1924
	str++;

1925 1926
	if (str[0] == '0' && str[1])
		return -EINVAL;
1927 1928 1929 1930 1931 1932 1933 1934 1935 1936 1937 1938
	len = _parse_integer(str, 16, &eval);
	if (len & KSTRTOX_OVERFLOW)
		return -EINVAL;
	if (eval != (unsigned long)eval)
		return -EINVAL;
	str += len;

	if (*str != '\0')
		return -EINVAL;

	*start = sval;
	*end = eval;
1939 1940 1941 1942

	return 0;
}

1943
static int map_files_d_revalidate(struct dentry *dentry, unsigned int flags)
1944 1945 1946 1947 1948 1949 1950 1951
{
	unsigned long vm_start, vm_end;
	bool exact_vma_exists = false;
	struct mm_struct *mm = NULL;
	struct task_struct *task;
	struct inode *inode;
	int status = 0;

1952
	if (flags & LOOKUP_RCU)
1953 1954
		return -ECHILD;

1955
	inode = d_inode(dentry);
1956 1957 1958 1959
	task = get_proc_task(inode);
	if (!task)
		goto out_notask;

1960
	mm = mm_access(task, PTRACE_MODE_READ_FSCREDS);
1961
	if (IS_ERR_OR_NULL(mm))
1962 1963 1964 1965 1966 1967 1968 1969 1970 1971 1972
		goto out;

	if (!dname_to_vma_addr(dentry, &vm_start, &vm_end)) {
		down_read(&mm->mmap_sem);
		exact_vma_exists = !!find_exact_vma(mm, vm_start, vm_end);
		up_read(&mm->mmap_sem);
	}

	mmput(mm);

	if (exact_vma_exists) {
1973 1974
		task_dump_owner(task, 0, &inode->i_uid, &inode->i_gid);

1975 1976 1977 1978 1979 1980 1981 1982 1983 1984 1985 1986 1987 1988 1989 1990
		security_task_to_inode(task, inode);
		status = 1;
	}

out:
	put_task_struct(task);

out_notask:
	return status;
}

static const struct dentry_operations tid_map_files_dentry_operations = {
	.d_revalidate	= map_files_d_revalidate,
	.d_delete	= pid_delete_dentry,
};

1991
static int map_files_get_link(struct dentry *dentry, struct path *path)
1992 1993 1994 1995 1996 1997 1998 1999
{
	unsigned long vm_start, vm_end;
	struct vm_area_struct *vma;
	struct task_struct *task;
	struct mm_struct *mm;
	int rc;

	rc = -ENOENT;
2000
	task = get_proc_task(d_inode(dentry));
2001 2002 2003 2004 2005 2006 2007 2008 2009 2010 2011 2012
	if (!task)
		goto out;

	mm = get_task_mm(task);
	put_task_struct(task);
	if (!mm)
		goto out;

	rc = dname_to_vma_addr(dentry, &vm_start, &vm_end);
	if (rc)
		goto out_mmput;

2013
	rc = -ENOENT;
2014 2015 2016 2017 2018 2019 2020 2021 2022 2023 2024 2025 2026 2027 2028 2029
	down_read(&mm->mmap_sem);
	vma = find_exact_vma(mm, vm_start, vm_end);
	if (vma && vma->vm_file) {
		*path = vma->vm_file->f_path;
		path_get(path);
		rc = 0;
	}
	up_read(&mm->mmap_sem);

out_mmput:
	mmput(mm);
out:
	return rc;
}

struct map_files_info {
2030 2031
	unsigned long	start;
	unsigned long	end;
2032
	fmode_t		mode;
2033 2034
};

2035 2036 2037 2038 2039 2040
/*
 * Only allow CAP_SYS_ADMIN to follow the links, due to concerns about how the
 * symlinks may be used to bypass permissions on ancestor directories in the
 * path to the file in question.
 */
static const char *
2041
proc_map_files_get_link(struct dentry *dentry,
2042 2043
			struct inode *inode,
		        struct delayed_call *done)
2044 2045 2046 2047
{
	if (!capable(CAP_SYS_ADMIN))
		return ERR_PTR(-EPERM);

2048
	return proc_pid_get_link(dentry, inode, done);
2049 2050 2051
}

/*
2052
 * Identical to proc_pid_link_inode_operations except for get_link()
2053 2054 2055
 */
static const struct inode_operations proc_map_files_link_inode_operations = {
	.readlink	= proc_pid_readlink,
2056
	.get_link	= proc_map_files_get_link,
2057 2058 2059
	.setattr	= proc_setattr,
};

2060 2061
static struct dentry *
proc_map_files_instantiate(struct dentry *dentry,
2062 2063
			   struct task_struct *task, const void *ptr)
{
2064
	fmode_t mode = (fmode_t)(unsigned long)ptr;
2065 2066 2067
	struct proc_inode *ei;
	struct inode *inode;

2068
	inode = proc_pid_make_inode(dentry->d_sb, task, S_IFLNK |
2069 2070
				    ((mode & FMODE_READ ) ? S_IRUSR : 0) |
				    ((mode & FMODE_WRITE) ? S_IWUSR : 0));
2071
	if (!inode)
2072
		return ERR_PTR(-ENOENT);
2073 2074

	ei = PROC_I(inode);
2075
	ei->op.proc_get_link = map_files_get_link;
2076

2077
	inode->i_op = &proc_map_files_link_inode_operations;
2078 2079 2080
	inode->i_size = 64;

	d_set_d_op(dentry, &tid_map_files_dentry_operations);
2081
	return d_splice_alias(inode, dentry);
2082 2083 2084
}

static struct dentry *proc_map_files_lookup(struct inode *dir,
A
Al Viro 已提交
2085
		struct dentry *dentry, unsigned int flags)
2086 2087 2088 2089
{
	unsigned long vm_start, vm_end;
	struct vm_area_struct *vma;
	struct task_struct *task;
2090
	struct dentry *result;
2091 2092
	struct mm_struct *mm;

2093
	result = ERR_PTR(-ENOENT);
2094 2095 2096 2097
	task = get_proc_task(dir);
	if (!task)
		goto out;

2098
	result = ERR_PTR(-EACCES);
2099
	if (!ptrace_may_access(task, PTRACE_MODE_READ_FSCREDS))
2100 2101
		goto out_put_task;

2102
	result = ERR_PTR(-ENOENT);
2103
	if (dname_to_vma_addr(dentry, &vm_start, &vm_end))
2104
		goto out_put_task;
2105 2106 2107

	mm = get_task_mm(task);
	if (!mm)
2108
		goto out_put_task;
2109 2110 2111 2112 2113 2114

	down_read(&mm->mmap_sem);
	vma = find_exact_vma(mm, vm_start, vm_end);
	if (!vma)
		goto out_no_vma;

2115
	if (vma->vm_file)
2116
		result = proc_map_files_instantiate(dentry, task,
2117
				(void *)(unsigned long)vma->vm_file->f_mode);
2118 2119 2120 2121 2122 2123 2124

out_no_vma:
	up_read(&mm->mmap_sem);
	mmput(mm);
out_put_task:
	put_task_struct(task);
out:
2125
	return result;
2126 2127 2128 2129 2130 2131 2132 2133 2134
}

static const struct inode_operations proc_map_files_inode_operations = {
	.lookup		= proc_map_files_lookup,
	.permission	= proc_fd_permission,
	.setattr	= proc_setattr,
};

static int
A
Al Viro 已提交
2135
proc_map_files_readdir(struct file *file, struct dir_context *ctx)
2136 2137 2138 2139
{
	struct vm_area_struct *vma;
	struct task_struct *task;
	struct mm_struct *mm;
A
Al Viro 已提交
2140 2141 2142 2143
	unsigned long nr_files, pos, i;
	struct flex_array *fa = NULL;
	struct map_files_info info;
	struct map_files_info *p;
2144 2145 2146
	int ret;

	ret = -ENOENT;
A
Al Viro 已提交
2147
	task = get_proc_task(file_inode(file));
2148 2149 2150 2151
	if (!task)
		goto out;

	ret = -EACCES;
2152
	if (!ptrace_may_access(task, PTRACE_MODE_READ_FSCREDS))
2153 2154 2155
		goto out_put_task;

	ret = 0;
A
Al Viro 已提交
2156 2157
	if (!dir_emit_dots(file, ctx))
		goto out_put_task;
2158

A
Al Viro 已提交
2159 2160 2161 2162
	mm = get_task_mm(task);
	if (!mm)
		goto out_put_task;
	down_read(&mm->mmap_sem);
2163

A
Al Viro 已提交
2164
	nr_files = 0;
2165

A
Al Viro 已提交
2166 2167 2168 2169 2170 2171 2172 2173 2174
	/*
	 * We need two passes here:
	 *
	 *  1) Collect vmas of mapped files with mmap_sem taken
	 *  2) Release mmap_sem and instantiate entries
	 *
	 * otherwise we get lockdep complained, since filldir()
	 * routine might require mmap_sem taken in might_fault().
	 */
2175

A
Al Viro 已提交
2176 2177 2178 2179 2180 2181 2182 2183 2184 2185 2186 2187 2188 2189 2190 2191
	for (vma = mm->mmap, pos = 2; vma; vma = vma->vm_next) {
		if (vma->vm_file && ++pos > ctx->pos)
			nr_files++;
	}

	if (nr_files) {
		fa = flex_array_alloc(sizeof(info), nr_files,
					GFP_KERNEL);
		if (!fa || flex_array_prealloc(fa, 0, nr_files,
						GFP_KERNEL)) {
			ret = -ENOMEM;
			if (fa)
				flex_array_free(fa);
			up_read(&mm->mmap_sem);
			mmput(mm);
			goto out_put_task;
2192
		}
A
Al Viro 已提交
2193 2194 2195 2196 2197 2198 2199
		for (i = 0, vma = mm->mmap, pos = 2; vma;
				vma = vma->vm_next) {
			if (!vma->vm_file)
				continue;
			if (++pos <= ctx->pos)
				continue;

2200 2201
			info.start = vma->vm_start;
			info.end = vma->vm_end;
A
Al Viro 已提交
2202 2203 2204
			info.mode = vma->vm_file->f_mode;
			if (flex_array_put(fa, i++, &info, GFP_KERNEL))
				BUG();
2205 2206
		}
	}
A
Al Viro 已提交
2207
	up_read(&mm->mmap_sem);
2208
	mmput(mm);
A
Al Viro 已提交
2209 2210

	for (i = 0; i < nr_files; i++) {
2211 2212 2213
		char buf[4 * sizeof(long) + 2];	/* max: %lx-%lx\0 */
		unsigned int len;

A
Al Viro 已提交
2214
		p = flex_array_get(fa, i);
2215
		len = snprintf(buf, sizeof(buf), "%lx-%lx", p->start, p->end);
A
Al Viro 已提交
2216
		if (!proc_fill_cache(file, ctx,
2217
				      buf, len,
A
Al Viro 已提交
2218 2219 2220 2221 2222
				      proc_map_files_instantiate,
				      task,
				      (void *)(unsigned long)p->mode))
			break;
		ctx->pos++;
2223
	}
A
Al Viro 已提交
2224 2225
	if (fa)
		flex_array_free(fa);
2226 2227 2228 2229 2230 2231 2232 2233 2234

out_put_task:
	put_task_struct(task);
out:
	return ret;
}

static const struct file_operations proc_map_files_operations = {
	.read		= generic_read_dir,
2235 2236
	.iterate_shared	= proc_map_files_readdir,
	.llseek		= generic_file_llseek,
2237 2238
};

2239
#if defined(CONFIG_CHECKPOINT_RESTORE) && defined(CONFIG_POSIX_TIMERS)
2240 2241 2242 2243
struct timers_private {
	struct pid *pid;
	struct task_struct *task;
	struct sighand_struct *sighand;
2244
	struct pid_namespace *ns;
2245 2246 2247 2248 2249 2250 2251 2252 2253 2254 2255 2256 2257 2258 2259 2260 2261 2262 2263 2264 2265 2266 2267 2268 2269 2270 2271 2272 2273 2274 2275 2276 2277 2278 2279 2280 2281 2282 2283 2284 2285 2286
	unsigned long flags;
};

static void *timers_start(struct seq_file *m, loff_t *pos)
{
	struct timers_private *tp = m->private;

	tp->task = get_pid_task(tp->pid, PIDTYPE_PID);
	if (!tp->task)
		return ERR_PTR(-ESRCH);

	tp->sighand = lock_task_sighand(tp->task, &tp->flags);
	if (!tp->sighand)
		return ERR_PTR(-ESRCH);

	return seq_list_start(&tp->task->signal->posix_timers, *pos);
}

static void *timers_next(struct seq_file *m, void *v, loff_t *pos)
{
	struct timers_private *tp = m->private;
	return seq_list_next(v, &tp->task->signal->posix_timers, pos);
}

static void timers_stop(struct seq_file *m, void *v)
{
	struct timers_private *tp = m->private;

	if (tp->sighand) {
		unlock_task_sighand(tp->task, &tp->flags);
		tp->sighand = NULL;
	}

	if (tp->task) {
		put_task_struct(tp->task);
		tp->task = NULL;
	}
}

static int show_timer(struct seq_file *m, void *v)
{
	struct k_itimer *timer;
2287 2288
	struct timers_private *tp = m->private;
	int notify;
2289
	static const char * const nstr[] = {
2290 2291 2292 2293
		[SIGEV_SIGNAL] = "signal",
		[SIGEV_NONE] = "none",
		[SIGEV_THREAD] = "thread",
	};
2294 2295

	timer = list_entry((struct list_head *)v, struct k_itimer, list);
2296 2297
	notify = timer->it_sigev_notify;

2298
	seq_printf(m, "ID: %d\n", timer->it_id);
2299
	seq_printf(m, "signal: %d/%px\n",
2300 2301
		   timer->sigq->info.si_signo,
		   timer->sigq->info.si_value.sival_ptr);
2302
	seq_printf(m, "notify: %s/%s.%d\n",
2303 2304 2305
		   nstr[notify & ~SIGEV_THREAD_ID],
		   (notify & SIGEV_THREAD_ID) ? "tid" : "pid",
		   pid_nr_ns(timer->it_pid, tp->ns));
2306
	seq_printf(m, "ClockID: %d\n", timer->it_clock);
2307 2308 2309 2310 2311 2312 2313 2314 2315 2316 2317 2318 2319 2320 2321 2322 2323 2324 2325 2326 2327

	return 0;
}

static const struct seq_operations proc_timers_seq_ops = {
	.start	= timers_start,
	.next	= timers_next,
	.stop	= timers_stop,
	.show	= show_timer,
};

static int proc_timers_open(struct inode *inode, struct file *file)
{
	struct timers_private *tp;

	tp = __seq_open_private(file, &proc_timers_seq_ops,
			sizeof(struct timers_private));
	if (!tp)
		return -ENOMEM;

	tp->pid = proc_pid(inode);
2328
	tp->ns = proc_pid_ns(inode);
2329 2330 2331 2332 2333 2334 2335 2336 2337
	return 0;
}

static const struct file_operations proc_timers_operations = {
	.open		= proc_timers_open,
	.read		= seq_read,
	.llseek		= seq_lseek,
	.release	= seq_release_private,
};
2338
#endif
2339

2340 2341 2342 2343 2344 2345 2346 2347 2348 2349 2350 2351 2352 2353 2354 2355
static ssize_t timerslack_ns_write(struct file *file, const char __user *buf,
					size_t count, loff_t *offset)
{
	struct inode *inode = file_inode(file);
	struct task_struct *p;
	u64 slack_ns;
	int err;

	err = kstrtoull_from_user(buf, count, 10, &slack_ns);
	if (err < 0)
		return err;

	p = get_proc_task(inode);
	if (!p)
		return -ESRCH;

2356
	if (p != current) {
2357 2358 2359
		rcu_read_lock();
		if (!ns_capable(__task_cred(p)->user_ns, CAP_SYS_NICE)) {
			rcu_read_unlock();
2360 2361 2362
			count = -EPERM;
			goto out;
		}
2363
		rcu_read_unlock();
2364

2365 2366 2367 2368 2369
		err = security_task_setscheduler(p);
		if (err) {
			count = err;
			goto out;
		}
2370 2371
	}

2372 2373 2374 2375 2376 2377 2378 2379
	task_lock(p);
	if (slack_ns == 0)
		p->timer_slack_ns = p->default_timer_slack_ns;
	else
		p->timer_slack_ns = slack_ns;
	task_unlock(p);

out:
2380 2381 2382 2383 2384 2385 2386 2387 2388
	put_task_struct(p);

	return count;
}

static int timerslack_ns_show(struct seq_file *m, void *v)
{
	struct inode *inode = m->private;
	struct task_struct *p;
2389
	int err = 0;
2390 2391 2392 2393 2394

	p = get_proc_task(inode);
	if (!p)
		return -ESRCH;

2395
	if (p != current) {
2396 2397 2398
		rcu_read_lock();
		if (!ns_capable(__task_cred(p)->user_ns, CAP_SYS_NICE)) {
			rcu_read_unlock();
2399 2400 2401
			err = -EPERM;
			goto out;
		}
2402 2403
		rcu_read_unlock();

2404 2405 2406 2407
		err = security_task_getscheduler(p);
		if (err)
			goto out;
	}
2408

2409 2410 2411 2412 2413
	task_lock(p);
	seq_printf(m, "%llu\n", p->timer_slack_ns);
	task_unlock(p);

out:
2414 2415 2416 2417 2418 2419 2420 2421 2422 2423 2424 2425 2426 2427 2428 2429 2430 2431
	put_task_struct(p);

	return err;
}

static int timerslack_ns_open(struct inode *inode, struct file *filp)
{
	return single_open(filp, timerslack_ns_show, inode);
}

static const struct file_operations proc_pid_set_timerslack_ns_operations = {
	.open		= timerslack_ns_open,
	.read		= seq_read,
	.write		= timerslack_ns_write,
	.llseek		= seq_lseek,
	.release	= single_release,
};

2432 2433
static struct dentry *proc_pident_instantiate(struct dentry *dentry,
	struct task_struct *task, const void *ptr)
2434
{
2435
	const struct pid_entry *p = ptr;
2436 2437 2438
	struct inode *inode;
	struct proc_inode *ei;

2439
	inode = proc_pid_make_inode(dentry->d_sb, task, p->mode);
2440
	if (!inode)
2441
		return ERR_PTR(-ENOENT);
2442 2443 2444

	ei = PROC_I(inode);
	if (S_ISDIR(inode->i_mode))
M
Miklos Szeredi 已提交
2445
		set_nlink(inode, 2);	/* Use getattr to fix if necessary */
2446 2447 2448 2449 2450
	if (p->iop)
		inode->i_op = p->iop;
	if (p->fop)
		inode->i_fop = p->fop;
	ei->op = p->op;
2451
	pid_update_inode(task, inode);
2452
	d_set_d_op(dentry, &pid_dentry_operations);
2453
	return d_splice_alias(inode, dentry);
2454 2455
}

L
Linus Torvalds 已提交
2456 2457
static struct dentry *proc_pident_lookup(struct inode *dir, 
					 struct dentry *dentry,
2458
					 const struct pid_entry *ents,
2459
					 unsigned int nents)
L
Linus Torvalds 已提交
2460
{
2461
	struct task_struct *task = get_proc_task(dir);
2462
	const struct pid_entry *p, *last;
2463
	struct dentry *res = ERR_PTR(-ENOENT);
L
Linus Torvalds 已提交
2464

2465 2466
	if (!task)
		goto out_no_task;
L
Linus Torvalds 已提交
2467

2468 2469 2470 2471
	/*
	 * Yes, it does not scale. And it should not. Don't add
	 * new entries into /proc/<tgid>/ without very good reasons.
	 */
2472 2473
	last = &ents[nents];
	for (p = ents; p < last; p++) {
L
Linus Torvalds 已提交
2474 2475
		if (p->len != dentry->d_name.len)
			continue;
2476 2477
		if (!memcmp(dentry->d_name.name, p->name, p->len)) {
			res = proc_pident_instantiate(dentry, task, p);
L
Linus Torvalds 已提交
2478
			break;
2479
		}
L
Linus Torvalds 已提交
2480
	}
2481 2482
	put_task_struct(task);
out_no_task:
2483
	return res;
L
Linus Torvalds 已提交
2484 2485
}

A
Al Viro 已提交
2486
static int proc_pident_readdir(struct file *file, struct dir_context *ctx,
2487
		const struct pid_entry *ents, unsigned int nents)
2488
{
A
Al Viro 已提交
2489 2490
	struct task_struct *task = get_proc_task(file_inode(file));
	const struct pid_entry *p;
2491 2492

	if (!task)
A
Al Viro 已提交
2493
		return -ENOENT;
2494

A
Al Viro 已提交
2495 2496 2497 2498 2499
	if (!dir_emit_dots(file, ctx))
		goto out;

	if (ctx->pos >= nents + 2)
		goto out;
2500

2501
	for (p = ents + (ctx->pos - 2); p < ents + nents; p++) {
A
Al Viro 已提交
2502 2503 2504 2505 2506
		if (!proc_fill_cache(file, ctx, p->name, p->len,
				proc_pident_instantiate, task, p))
			break;
		ctx->pos++;
	}
2507
out:
2508
	put_task_struct(task);
A
Al Viro 已提交
2509
	return 0;
L
Linus Torvalds 已提交
2510 2511
}

2512 2513 2514 2515
#ifdef CONFIG_SECURITY
static ssize_t proc_pid_attr_read(struct file * file, char __user * buf,
				  size_t count, loff_t *ppos)
{
A
Al Viro 已提交
2516
	struct inode * inode = file_inode(file);
2517
	char *p = NULL;
2518 2519 2520 2521
	ssize_t length;
	struct task_struct *task = get_proc_task(inode);

	if (!task)
2522
		return -ESRCH;
2523 2524

	length = security_getprocattr(task,
2525
				      (char*)file->f_path.dentry->d_name.name,
2526
				      &p);
2527
	put_task_struct(task);
2528 2529 2530
	if (length > 0)
		length = simple_read_from_buffer(buf, count, ppos, p, length);
	kfree(p);
2531
	return length;
L
Linus Torvalds 已提交
2532 2533
}

2534 2535 2536
static ssize_t proc_pid_attr_write(struct file * file, const char __user * buf,
				   size_t count, loff_t *ppos)
{
A
Al Viro 已提交
2537
	struct inode * inode = file_inode(file);
2538
	struct task_struct *task;
2539
	void *page;
2540
	int rv;
2541

2542 2543 2544 2545 2546 2547
	rcu_read_lock();
	task = pid_task(proc_pid(inode), PIDTYPE_PID);
	if (!task) {
		rcu_read_unlock();
		return -ESRCH;
	}
2548
	/* A task may only write its own attributes. */
2549 2550 2551 2552 2553
	if (current != task) {
		rcu_read_unlock();
		return -EACCES;
	}
	rcu_read_unlock();
2554

2555 2556 2557 2558 2559
	if (count > PAGE_SIZE)
		count = PAGE_SIZE;

	/* No partial writes. */
	if (*ppos != 0)
2560
		return -EINVAL;
2561

2562 2563
	page = memdup_user(buf, count);
	if (IS_ERR(page)) {
2564
		rv = PTR_ERR(page);
2565
		goto out;
2566
	}
2567

2568
	/* Guard against adverse ptrace interaction */
2569 2570
	rv = mutex_lock_interruptible(&current->signal->cred_guard_mutex);
	if (rv < 0)
2571 2572
		goto out_free;

2573
	rv = security_setprocattr(file->f_path.dentry->d_name.name, page, count);
2574
	mutex_unlock(&current->signal->cred_guard_mutex);
2575
out_free:
2576
	kfree(page);
2577
out:
2578
	return rv;
2579 2580
}

2581
static const struct file_operations proc_pid_attr_operations = {
2582 2583
	.read		= proc_pid_attr_read,
	.write		= proc_pid_attr_write,
2584
	.llseek		= generic_file_llseek,
2585 2586
};

2587
static const struct pid_entry attr_dir_stuff[] = {
A
Alexey Dobriyan 已提交
2588 2589 2590 2591 2592 2593
	REG("current",    S_IRUGO|S_IWUGO, proc_pid_attr_operations),
	REG("prev",       S_IRUGO,	   proc_pid_attr_operations),
	REG("exec",       S_IRUGO|S_IWUGO, proc_pid_attr_operations),
	REG("fscreate",   S_IRUGO|S_IWUGO, proc_pid_attr_operations),
	REG("keycreate",  S_IRUGO|S_IWUGO, proc_pid_attr_operations),
	REG("sockcreate", S_IRUGO|S_IWUGO, proc_pid_attr_operations),
2594 2595
};

A
Al Viro 已提交
2596
static int proc_attr_dir_readdir(struct file *file, struct dir_context *ctx)
2597
{
A
Al Viro 已提交
2598 2599
	return proc_pident_readdir(file, ctx, 
				   attr_dir_stuff, ARRAY_SIZE(attr_dir_stuff));
2600 2601
}

2602
static const struct file_operations proc_attr_dir_operations = {
L
Linus Torvalds 已提交
2603
	.read		= generic_read_dir,
2604 2605
	.iterate_shared	= proc_attr_dir_readdir,
	.llseek		= generic_file_llseek,
L
Linus Torvalds 已提交
2606 2607
};

2608
static struct dentry *proc_attr_dir_lookup(struct inode *dir,
A
Al Viro 已提交
2609
				struct dentry *dentry, unsigned int flags)
2610
{
2611 2612
	return proc_pident_lookup(dir, dentry,
				  attr_dir_stuff, ARRAY_SIZE(attr_dir_stuff));
2613 2614
}

2615
static const struct inode_operations proc_attr_dir_inode_operations = {
2616
	.lookup		= proc_attr_dir_lookup,
2617
	.getattr	= pid_getattr,
2618
	.setattr	= proc_setattr,
L
Linus Torvalds 已提交
2619 2620
};

2621 2622
#endif

C
Christoph Hellwig 已提交
2623
#ifdef CONFIG_ELF_CORE
2624 2625 2626
static ssize_t proc_coredump_filter_read(struct file *file, char __user *buf,
					 size_t count, loff_t *ppos)
{
A
Al Viro 已提交
2627
	struct task_struct *task = get_proc_task(file_inode(file));
2628 2629 2630 2631 2632 2633 2634 2635 2636 2637 2638 2639 2640 2641 2642 2643 2644 2645 2646 2647 2648 2649 2650 2651 2652 2653 2654 2655 2656 2657 2658 2659 2660 2661 2662
	struct mm_struct *mm;
	char buffer[PROC_NUMBUF];
	size_t len;
	int ret;

	if (!task)
		return -ESRCH;

	ret = 0;
	mm = get_task_mm(task);
	if (mm) {
		len = snprintf(buffer, sizeof(buffer), "%08lx\n",
			       ((mm->flags & MMF_DUMP_FILTER_MASK) >>
				MMF_DUMP_FILTER_SHIFT));
		mmput(mm);
		ret = simple_read_from_buffer(buf, count, ppos, buffer, len);
	}

	put_task_struct(task);

	return ret;
}

static ssize_t proc_coredump_filter_write(struct file *file,
					  const char __user *buf,
					  size_t count,
					  loff_t *ppos)
{
	struct task_struct *task;
	struct mm_struct *mm;
	unsigned int val;
	int ret;
	int i;
	unsigned long mask;

2663 2664 2665
	ret = kstrtouint_from_user(buf, count, 0, &val);
	if (ret < 0)
		return ret;
2666 2667

	ret = -ESRCH;
A
Al Viro 已提交
2668
	task = get_proc_task(file_inode(file));
2669 2670 2671 2672 2673 2674
	if (!task)
		goto out_no_task;

	mm = get_task_mm(task);
	if (!mm)
		goto out_no_mm;
2675
	ret = 0;
2676 2677 2678 2679 2680 2681 2682 2683 2684 2685 2686 2687

	for (i = 0, mask = 1; i < MMF_DUMP_FILTER_BITS; i++, mask <<= 1) {
		if (val & mask)
			set_bit(i + MMF_DUMP_FILTER_SHIFT, &mm->flags);
		else
			clear_bit(i + MMF_DUMP_FILTER_SHIFT, &mm->flags);
	}

	mmput(mm);
 out_no_mm:
	put_task_struct(task);
 out_no_task:
2688 2689 2690
	if (ret < 0)
		return ret;
	return count;
2691 2692 2693 2694 2695
}

static const struct file_operations proc_coredump_filter_operations = {
	.read		= proc_coredump_filter_read,
	.write		= proc_coredump_filter_write,
2696
	.llseek		= generic_file_llseek,
2697 2698 2699
};
#endif

2700
#ifdef CONFIG_TASK_IO_ACCOUNTING
2701
static int do_io_accounting(struct task_struct *task, struct seq_file *m, int whole)
2702
{
2703
	struct task_io_accounting acct = task->ioac;
2704
	unsigned long flags;
2705
	int result;
2706

2707 2708 2709 2710
	result = mutex_lock_killable(&task->signal->cred_guard_mutex);
	if (result)
		return result;

2711
	if (!ptrace_may_access(task, PTRACE_MODE_READ_FSCREDS)) {
2712 2713 2714
		result = -EACCES;
		goto out_unlock;
	}
2715

2716 2717 2718 2719 2720 2721 2722 2723
	if (whole && lock_task_sighand(task, &flags)) {
		struct task_struct *t = task;

		task_io_accounting_add(&acct, &task->signal->ioac);
		while_each_thread(task, t)
			task_io_accounting_add(&acct, &t->ioac);

		unlock_task_sighand(task, &flags);
2724
	}
2725 2726 2727 2728 2729 2730 2731 2732 2733 2734 2735 2736 2737 2738 2739 2740 2741
	seq_printf(m,
		   "rchar: %llu\n"
		   "wchar: %llu\n"
		   "syscr: %llu\n"
		   "syscw: %llu\n"
		   "read_bytes: %llu\n"
		   "write_bytes: %llu\n"
		   "cancelled_write_bytes: %llu\n",
		   (unsigned long long)acct.rchar,
		   (unsigned long long)acct.wchar,
		   (unsigned long long)acct.syscr,
		   (unsigned long long)acct.syscw,
		   (unsigned long long)acct.read_bytes,
		   (unsigned long long)acct.write_bytes,
		   (unsigned long long)acct.cancelled_write_bytes);
	result = 0;

2742 2743 2744
out_unlock:
	mutex_unlock(&task->signal->cred_guard_mutex);
	return result;
2745 2746
}

2747 2748
static int proc_tid_io_accounting(struct seq_file *m, struct pid_namespace *ns,
				  struct pid *pid, struct task_struct *task)
2749
{
2750
	return do_io_accounting(task, m, 0);
2751
}
2752

2753 2754
static int proc_tgid_io_accounting(struct seq_file *m, struct pid_namespace *ns,
				   struct pid *pid, struct task_struct *task)
2755
{
2756
	return do_io_accounting(task, m, 1);
2757 2758
}
#endif /* CONFIG_TASK_IO_ACCOUNTING */
2759

2760 2761
#ifdef CONFIG_USER_NS
static int proc_id_map_open(struct inode *inode, struct file *file,
F
Fabian Frederick 已提交
2762
	const struct seq_operations *seq_ops)
2763 2764 2765 2766 2767 2768 2769 2770 2771 2772 2773 2774 2775 2776 2777 2778 2779 2780 2781 2782 2783 2784 2785 2786 2787 2788 2789 2790 2791 2792 2793 2794 2795 2796 2797 2798 2799 2800 2801 2802 2803 2804 2805 2806 2807 2808 2809 2810
{
	struct user_namespace *ns = NULL;
	struct task_struct *task;
	struct seq_file *seq;
	int ret = -EINVAL;

	task = get_proc_task(inode);
	if (task) {
		rcu_read_lock();
		ns = get_user_ns(task_cred_xxx(task, user_ns));
		rcu_read_unlock();
		put_task_struct(task);
	}
	if (!ns)
		goto err;

	ret = seq_open(file, seq_ops);
	if (ret)
		goto err_put_ns;

	seq = file->private_data;
	seq->private = ns;

	return 0;
err_put_ns:
	put_user_ns(ns);
err:
	return ret;
}

static int proc_id_map_release(struct inode *inode, struct file *file)
{
	struct seq_file *seq = file->private_data;
	struct user_namespace *ns = seq->private;
	put_user_ns(ns);
	return seq_release(inode, file);
}

static int proc_uid_map_open(struct inode *inode, struct file *file)
{
	return proc_id_map_open(inode, file, &proc_uid_seq_operations);
}

static int proc_gid_map_open(struct inode *inode, struct file *file)
{
	return proc_id_map_open(inode, file, &proc_gid_seq_operations);
}

2811 2812 2813 2814 2815
static int proc_projid_map_open(struct inode *inode, struct file *file)
{
	return proc_id_map_open(inode, file, &proc_projid_seq_operations);
}

2816 2817 2818 2819 2820 2821 2822 2823 2824 2825 2826 2827 2828 2829 2830
static const struct file_operations proc_uid_map_operations = {
	.open		= proc_uid_map_open,
	.write		= proc_uid_map_write,
	.read		= seq_read,
	.llseek		= seq_lseek,
	.release	= proc_id_map_release,
};

static const struct file_operations proc_gid_map_operations = {
	.open		= proc_gid_map_open,
	.write		= proc_gid_map_write,
	.read		= seq_read,
	.llseek		= seq_lseek,
	.release	= proc_id_map_release,
};
2831 2832 2833 2834 2835 2836 2837 2838

static const struct file_operations proc_projid_map_operations = {
	.open		= proc_projid_map_open,
	.write		= proc_projid_map_write,
	.read		= seq_read,
	.llseek		= seq_lseek,
	.release	= proc_id_map_release,
};
2839 2840 2841 2842 2843 2844 2845 2846 2847 2848 2849 2850 2851 2852 2853 2854 2855 2856 2857 2858 2859 2860 2861 2862 2863 2864 2865 2866 2867 2868 2869 2870 2871 2872 2873 2874 2875 2876 2877 2878 2879 2880 2881 2882 2883 2884 2885 2886 2887 2888 2889

static int proc_setgroups_open(struct inode *inode, struct file *file)
{
	struct user_namespace *ns = NULL;
	struct task_struct *task;
	int ret;

	ret = -ESRCH;
	task = get_proc_task(inode);
	if (task) {
		rcu_read_lock();
		ns = get_user_ns(task_cred_xxx(task, user_ns));
		rcu_read_unlock();
		put_task_struct(task);
	}
	if (!ns)
		goto err;

	if (file->f_mode & FMODE_WRITE) {
		ret = -EACCES;
		if (!ns_capable(ns, CAP_SYS_ADMIN))
			goto err_put_ns;
	}

	ret = single_open(file, &proc_setgroups_show, ns);
	if (ret)
		goto err_put_ns;

	return 0;
err_put_ns:
	put_user_ns(ns);
err:
	return ret;
}

static int proc_setgroups_release(struct inode *inode, struct file *file)
{
	struct seq_file *seq = file->private_data;
	struct user_namespace *ns = seq->private;
	int ret = single_release(inode, file);
	put_user_ns(ns);
	return ret;
}

static const struct file_operations proc_setgroups_operations = {
	.open		= proc_setgroups_open,
	.write		= proc_setgroups_write,
	.read		= seq_read,
	.llseek		= seq_lseek,
	.release	= proc_setgroups_release,
};
2890 2891
#endif /* CONFIG_USER_NS */

2892 2893 2894
static int proc_pid_personality(struct seq_file *m, struct pid_namespace *ns,
				struct pid *pid, struct task_struct *task)
{
2895 2896 2897 2898 2899 2900
	int err = lock_trace(task);
	if (!err) {
		seq_printf(m, "%08x\n", task->personality);
		unlock_trace(task);
	}
	return err;
2901 2902
}

2903 2904 2905 2906 2907 2908 2909 2910 2911
#ifdef CONFIG_LIVEPATCH
static int proc_pid_patch_state(struct seq_file *m, struct pid_namespace *ns,
				struct pid *pid, struct task_struct *task)
{
	seq_printf(m, "%d\n", task->patch_state);
	return 0;
}
#endif /* CONFIG_LIVEPATCH */

2912 2913 2914 2915 2916 2917 2918 2919 2920 2921 2922 2923 2924 2925 2926
#ifdef CONFIG_STACKLEAK_METRICS
static int proc_stack_depth(struct seq_file *m, struct pid_namespace *ns,
				struct pid *pid, struct task_struct *task)
{
	unsigned long prev_depth = THREAD_SIZE -
				(task->prev_lowest_stack & (THREAD_SIZE - 1));
	unsigned long depth = THREAD_SIZE -
				(task->lowest_stack & (THREAD_SIZE - 1));

	seq_printf(m, "previous stack depth: %lu\nstack depth: %lu\n",
							prev_depth, depth);
	return 0;
}
#endif /* CONFIG_STACKLEAK_METRICS */

2927 2928 2929
/*
 * Thread groups
 */
2930
static const struct file_operations proc_task_operations;
2931
static const struct inode_operations proc_task_inode_operations;
2932

2933
static const struct pid_entry tgid_base_stuff[] = {
A
Alexey Dobriyan 已提交
2934 2935
	DIR("task",       S_IRUGO|S_IXUGO, proc_task_inode_operations, proc_task_operations),
	DIR("fd",         S_IRUSR|S_IXUSR, proc_fd_inode_operations, proc_fd_operations),
2936
	DIR("map_files",  S_IRUSR|S_IXUSR, proc_map_files_inode_operations, proc_map_files_operations),
A
Alexey Dobriyan 已提交
2937
	DIR("fdinfo",     S_IRUSR|S_IXUSR, proc_fdinfo_inode_operations, proc_fdinfo_operations),
2938
	DIR("ns",	  S_IRUSR|S_IXUGO, proc_ns_dir_inode_operations, proc_ns_dir_operations),
A
Andrew Morton 已提交
2939
#ifdef CONFIG_NET
A
Alexey Dobriyan 已提交
2940
	DIR("net",        S_IRUGO|S_IXUGO, proc_net_inode_operations, proc_net_operations),
A
Andrew Morton 已提交
2941
#endif
A
Alexey Dobriyan 已提交
2942
	REG("environ",    S_IRUSR, proc_environ_operations),
A
Al Viro 已提交
2943
	REG("auxv",       S_IRUSR, proc_auxv_operations),
A
Alexey Dobriyan 已提交
2944
	ONE("status",     S_IRUGO, proc_pid_status),
2945
	ONE("personality", S_IRUSR, proc_pid_personality),
2946
	ONE("limits",	  S_IRUGO, proc_pid_limits),
I
Ingo Molnar 已提交
2947
#ifdef CONFIG_SCHED_DEBUG
A
Alexey Dobriyan 已提交
2948
	REG("sched",      S_IRUGO|S_IWUSR, proc_pid_sched_operations),
2949 2950 2951
#endif
#ifdef CONFIG_SCHED_AUTOGROUP
	REG("autogroup",  S_IRUGO|S_IWUSR, proc_pid_sched_autogroup_operations),
R
Roland McGrath 已提交
2952
#endif
2953
	REG("comm",      S_IRUGO|S_IWUSR, proc_pid_set_comm_operations),
R
Roland McGrath 已提交
2954
#ifdef CONFIG_HAVE_ARCH_TRACEHOOK
2955
	ONE("syscall",    S_IRUSR, proc_pid_syscall),
I
Ingo Molnar 已提交
2956
#endif
2957
	REG("cmdline",    S_IRUGO, proc_pid_cmdline_ops),
A
Alexey Dobriyan 已提交
2958 2959
	ONE("stat",       S_IRUGO, proc_tgid_stat),
	ONE("statm",      S_IRUGO, proc_pid_statm),
2960
	REG("maps",       S_IRUGO, proc_pid_maps_operations),
2961
#ifdef CONFIG_NUMA
2962
	REG("numa_maps",  S_IRUGO, proc_pid_numa_maps_operations),
2963
#endif
A
Alexey Dobriyan 已提交
2964 2965 2966 2967 2968 2969 2970
	REG("mem",        S_IRUSR|S_IWUSR, proc_mem_operations),
	LNK("cwd",        proc_cwd_link),
	LNK("root",       proc_root_link),
	LNK("exe",        proc_exe_link),
	REG("mounts",     S_IRUGO, proc_mounts_operations),
	REG("mountinfo",  S_IRUGO, proc_mountinfo_operations),
	REG("mountstats", S_IRUSR, proc_mountstats_operations),
2971
#ifdef CONFIG_PROC_PAGE_MONITOR
A
Alexey Dobriyan 已提交
2972
	REG("clear_refs", S_IWUSR, proc_clear_refs_operations),
2973
	REG("smaps",      S_IRUGO, proc_pid_smaps_operations),
2974
	REG("smaps_rollup", S_IRUGO, proc_pid_smaps_rollup_operations),
2975
	REG("pagemap",    S_IRUSR, proc_pagemap_operations),
2976 2977
#endif
#ifdef CONFIG_SECURITY
A
Alexey Dobriyan 已提交
2978
	DIR("attr",       S_IRUGO|S_IXUGO, proc_attr_dir_inode_operations, proc_attr_dir_operations),
2979 2980
#endif
#ifdef CONFIG_KALLSYMS
2981
	ONE("wchan",      S_IRUGO, proc_pid_wchan),
2982
#endif
K
Ken Chen 已提交
2983
#ifdef CONFIG_STACKTRACE
2984
	ONE("stack",      S_IRUSR, proc_pid_stack),
2985
#endif
2986
#ifdef CONFIG_SCHED_INFO
2987
	ONE("schedstat",  S_IRUGO, proc_pid_schedstat),
2988
#endif
A
Arjan van de Ven 已提交
2989
#ifdef CONFIG_LATENCYTOP
A
Alexey Dobriyan 已提交
2990
	REG("latency",  S_IRUGO, proc_lstats_operations),
A
Arjan van de Ven 已提交
2991
#endif
2992
#ifdef CONFIG_PROC_PID_CPUSET
Z
Zefan Li 已提交
2993
	ONE("cpuset",     S_IRUGO, proc_cpuset_show),
2994 2995
#endif
#ifdef CONFIG_CGROUPS
Z
Zefan Li 已提交
2996
	ONE("cgroup",  S_IRUGO, proc_cgroup_show),
2997
#endif
2998
	ONE("oom_score",  S_IRUGO, proc_oom_score),
2999
	REG("oom_adj",    S_IRUGO|S_IWUSR, proc_oom_adj_operations),
D
David Rientjes 已提交
3000
	REG("oom_score_adj", S_IRUGO|S_IWUSR, proc_oom_score_adj_operations),
3001
#ifdef CONFIG_AUDITSYSCALL
A
Alexey Dobriyan 已提交
3002 3003
	REG("loginuid",   S_IWUSR|S_IRUGO, proc_loginuid_operations),
	REG("sessionid",  S_IRUGO, proc_sessionid_operations),
3004
#endif
3005
#ifdef CONFIG_FAULT_INJECTION
A
Alexey Dobriyan 已提交
3006
	REG("make-it-fail", S_IRUGO|S_IWUSR, proc_fault_inject_operations),
3007
	REG("fail-nth", 0644, proc_fail_nth_operations),
3008
#endif
C
Christoph Hellwig 已提交
3009
#ifdef CONFIG_ELF_CORE
A
Alexey Dobriyan 已提交
3010
	REG("coredump_filter", S_IRUGO|S_IWUSR, proc_coredump_filter_operations),
3011
#endif
3012
#ifdef CONFIG_TASK_IO_ACCOUNTING
3013
	ONE("io",	S_IRUSR, proc_tgid_io_accounting),
3014
#endif
3015 3016 3017
#ifdef CONFIG_USER_NS
	REG("uid_map",    S_IRUGO|S_IWUSR, proc_uid_map_operations),
	REG("gid_map",    S_IRUGO|S_IWUSR, proc_gid_map_operations),
3018
	REG("projid_map", S_IRUGO|S_IWUSR, proc_projid_map_operations),
3019
	REG("setgroups",  S_IRUGO|S_IWUSR, proc_setgroups_operations),
3020
#endif
3021
#if defined(CONFIG_CHECKPOINT_RESTORE) && defined(CONFIG_POSIX_TIMERS)
3022 3023
	REG("timers",	  S_IRUGO, proc_timers_operations),
#endif
3024
	REG("timerslack_ns", S_IRUGO|S_IWUGO, proc_pid_set_timerslack_ns_operations),
3025 3026 3027
#ifdef CONFIG_LIVEPATCH
	ONE("patch_state",  S_IRUSR, proc_pid_patch_state),
#endif
3028 3029 3030
#ifdef CONFIG_STACKLEAK_METRICS
	ONE("stack_depth", S_IRUGO, proc_stack_depth),
#endif
3031
};
L
Linus Torvalds 已提交
3032

A
Al Viro 已提交
3033
static int proc_tgid_base_readdir(struct file *file, struct dir_context *ctx)
L
Linus Torvalds 已提交
3034
{
A
Al Viro 已提交
3035 3036
	return proc_pident_readdir(file, ctx,
				   tgid_base_stuff, ARRAY_SIZE(tgid_base_stuff));
L
Linus Torvalds 已提交
3037 3038
}

3039
static const struct file_operations proc_tgid_base_operations = {
L
Linus Torvalds 已提交
3040
	.read		= generic_read_dir,
3041 3042
	.iterate_shared	= proc_tgid_base_readdir,
	.llseek		= generic_file_llseek,
L
Linus Torvalds 已提交
3043 3044
};

A
Al Viro 已提交
3045 3046
static struct dentry *proc_tgid_base_lookup(struct inode *dir, struct dentry *dentry, unsigned int flags)
{
3047 3048
	return proc_pident_lookup(dir, dentry,
				  tgid_base_stuff, ARRAY_SIZE(tgid_base_stuff));
L
Linus Torvalds 已提交
3049 3050
}

3051
static const struct inode_operations proc_tgid_base_inode_operations = {
3052
	.lookup		= proc_tgid_base_lookup,
3053
	.getattr	= pid_getattr,
3054
	.setattr	= proc_setattr,
3055
	.permission	= proc_pid_permission,
L
Linus Torvalds 已提交
3056 3057
};

3058
static void proc_flush_task_mnt(struct vfsmount *mnt, pid_t pid, pid_t tgid)
L
Linus Torvalds 已提交
3059
{
3060
	struct dentry *dentry, *leader, *dir;
3061
	char buf[10 + 1];
3062 3063 3064
	struct qstr name;

	name.name = buf;
3065
	name.len = snprintf(buf, sizeof(buf), "%u", pid);
3066
	/* no ->d_hash() rejects on procfs */
3067
	dentry = d_hash_and_lookup(mnt->mnt_root, &name);
3068
	if (dentry) {
3069
		d_invalidate(dentry);
3070 3071
		dput(dentry);
	}
L
Linus Torvalds 已提交
3072

3073 3074 3075
	if (pid == tgid)
		return;

3076
	name.name = buf;
3077
	name.len = snprintf(buf, sizeof(buf), "%u", tgid);
3078
	leader = d_hash_and_lookup(mnt->mnt_root, &name);
3079 3080
	if (!leader)
		goto out;
L
Linus Torvalds 已提交
3081

3082 3083 3084 3085 3086 3087 3088
	name.name = "task";
	name.len = strlen(name.name);
	dir = d_hash_and_lookup(leader, &name);
	if (!dir)
		goto out_put_leader;

	name.name = buf;
3089
	name.len = snprintf(buf, sizeof(buf), "%u", pid);
3090 3091
	dentry = d_hash_and_lookup(dir, &name);
	if (dentry) {
3092
		d_invalidate(dentry);
3093
		dput(dentry);
L
Linus Torvalds 已提交
3094
	}
3095 3096 3097 3098 3099 3100

	dput(dir);
out_put_leader:
	dput(leader);
out:
	return;
L
Linus Torvalds 已提交
3101 3102
}

3103 3104 3105 3106 3107
/**
 * proc_flush_task -  Remove dcache entries for @task from the /proc dcache.
 * @task: task that should be flushed.
 *
 * When flushing dentries from proc, one needs to flush them from global
3108
 * proc (proc_mnt) and from all the namespaces' procs this task was seen
3109 3110 3111 3112 3113 3114 3115 3116 3117 3118 3119 3120 3121 3122 3123 3124 3125
 * in. This call is supposed to do all of this job.
 *
 * Looks in the dcache for
 * /proc/@pid
 * /proc/@tgid/task/@pid
 * if either directory is present flushes it and all of it'ts children
 * from the dcache.
 *
 * It is safe and reasonable to cache /proc entries for a task until
 * that task exits.  After that they just clog up the dcache with
 * useless entries, possibly causing useful dcache entries to be
 * flushed instead.  This routine is proved to flush those useless
 * dcache entries at process exit time.
 *
 * NOTE: This routine is just an optimization so it does not guarantee
 *       that no dcache entries will exist at process exit time it
 *       just makes it very unlikely that any will persist.
3126 3127 3128 3129
 */

void proc_flush_task(struct task_struct *task)
{
3130
	int i;
3131
	struct pid *pid, *tgid;
3132 3133 3134
	struct upid *upid;

	pid = task_pid(task);
3135
	tgid = task_tgid(task);
3136

3137
	for (i = 0; i <= pid->level; i++) {
3138 3139
		upid = &pid->numbers[i];
		proc_flush_task_mnt(upid->ns->proc_mnt, upid->nr,
3140
					tgid->numbers[i].nr);
3141
	}
3142 3143
}

3144
static struct dentry *proc_pid_instantiate(struct dentry * dentry,
3145
				   struct task_struct *task, const void *ptr)
3146 3147 3148
{
	struct inode *inode;

3149
	inode = proc_pid_make_inode(dentry->d_sb, task, S_IFDIR | S_IRUGO | S_IXUGO);
3150
	if (!inode)
3151
		return ERR_PTR(-ENOENT);
3152 3153 3154 3155

	inode->i_op = &proc_tgid_base_inode_operations;
	inode->i_fop = &proc_tgid_base_operations;
	inode->i_flags|=S_IMMUTABLE;
3156

3157
	set_nlink(inode, nlink_tgid);
3158
	pid_update_inode(task, inode);
3159

3160
	d_set_d_op(dentry, &pid_dentry_operations);
3161
	return d_splice_alias(inode, dentry);
3162 3163
}

3164
struct dentry *proc_pid_lookup(struct dentry *dentry, unsigned int flags)
L
Linus Torvalds 已提交
3165 3166 3167
{
	struct task_struct *task;
	unsigned tgid;
3168
	struct pid_namespace *ns;
3169
	struct dentry *result = ERR_PTR(-ENOENT);
L
Linus Torvalds 已提交
3170

3171
	tgid = name_to_int(&dentry->d_name);
L
Linus Torvalds 已提交
3172 3173 3174
	if (tgid == ~0U)
		goto out;

3175
	ns = dentry->d_sb->s_fs_info;
3176
	rcu_read_lock();
3177
	task = find_task_by_pid_ns(tgid, ns);
L
Linus Torvalds 已提交
3178 3179
	if (task)
		get_task_struct(task);
3180
	rcu_read_unlock();
L
Linus Torvalds 已提交
3181 3182 3183
	if (!task)
		goto out;

3184
	result = proc_pid_instantiate(dentry, task, NULL);
L
Linus Torvalds 已提交
3185 3186
	put_task_struct(task);
out:
3187
	return result;
L
Linus Torvalds 已提交
3188 3189 3190
}

/*
3191
 * Find the first task with tgid >= tgid
3192
 *
L
Linus Torvalds 已提交
3193
 */
3194 3195
struct tgid_iter {
	unsigned int tgid;
3196
	struct task_struct *task;
3197 3198 3199
};
static struct tgid_iter next_tgid(struct pid_namespace *ns, struct tgid_iter iter)
{
3200
	struct pid *pid;
L
Linus Torvalds 已提交
3201

3202 3203
	if (iter.task)
		put_task_struct(iter.task);
3204
	rcu_read_lock();
3205
retry:
3206 3207
	iter.task = NULL;
	pid = find_ge_pid(iter.tgid, ns);
3208
	if (pid) {
3209 3210
		iter.tgid = pid_nr_ns(pid, ns);
		iter.task = pid_task(pid, PIDTYPE_PID);
3211 3212 3213 3214 3215 3216 3217 3218 3219 3220 3221 3222
		/* What we to know is if the pid we have find is the
		 * pid of a thread_group_leader.  Testing for task
		 * being a thread_group_leader is the obvious thing
		 * todo but there is a window when it fails, due to
		 * the pid transfer logic in de_thread.
		 *
		 * So we perform the straight forward test of seeing
		 * if the pid we have found is the pid of a thread
		 * group leader, and don't worry if the task we have
		 * found doesn't happen to be a thread group leader.
		 * As we don't care in the case of readdir.
		 */
3223 3224
		if (!iter.task || !has_group_leader_pid(iter.task)) {
			iter.tgid += 1;
3225
			goto retry;
3226 3227
		}
		get_task_struct(iter.task);
3228
	}
3229
	rcu_read_unlock();
3230
	return iter;
L
Linus Torvalds 已提交
3231 3232
}

3233
#define TGID_OFFSET (FIRST_PROCESS_ENTRY + 2)
3234

L
Linus Torvalds 已提交
3235
/* for the /proc/ directory itself, after non-process stuff has been done */
A
Al Viro 已提交
3236
int proc_pid_readdir(struct file *file, struct dir_context *ctx)
L
Linus Torvalds 已提交
3237
{
3238
	struct tgid_iter iter;
3239
	struct pid_namespace *ns = proc_pid_ns(file_inode(file));
A
Al Viro 已提交
3240
	loff_t pos = ctx->pos;
L
Linus Torvalds 已提交
3241

3242
	if (pos >= PID_MAX_LIMIT + TGID_OFFSET)
A
Al Viro 已提交
3243
		return 0;
L
Linus Torvalds 已提交
3244

3245
	if (pos == TGID_OFFSET - 2) {
3246
		struct inode *inode = d_inode(ns->proc_self);
3247
		if (!dir_emit(ctx, "self", 4, inode->i_ino, DT_LNK))
A
Al Viro 已提交
3248
			return 0;
3249 3250 3251
		ctx->pos = pos = pos + 1;
	}
	if (pos == TGID_OFFSET - 1) {
3252
		struct inode *inode = d_inode(ns->proc_thread_self);
3253 3254 3255
		if (!dir_emit(ctx, "thread-self", 11, inode->i_ino, DT_LNK))
			return 0;
		ctx->pos = pos = pos + 1;
3256
	}
3257
	iter.tgid = pos - TGID_OFFSET;
3258 3259 3260 3261
	iter.task = NULL;
	for (iter = next_tgid(ns, iter);
	     iter.task;
	     iter.tgid += 1, iter = next_tgid(ns, iter)) {
3262
		char name[10 + 1];
3263
		unsigned int len;
3264 3265

		cond_resched();
3266
		if (!has_pid_permissions(ns, iter.task, HIDEPID_INVISIBLE))
A
Al Viro 已提交
3267
			continue;
3268

3269
		len = snprintf(name, sizeof(name), "%u", iter.tgid);
A
Al Viro 已提交
3270 3271 3272
		ctx->pos = iter.tgid + TGID_OFFSET;
		if (!proc_fill_cache(file, ctx, name, len,
				     proc_pid_instantiate, iter.task, NULL)) {
3273
			put_task_struct(iter.task);
A
Al Viro 已提交
3274
			return 0;
L
Linus Torvalds 已提交
3275
		}
3276
	}
A
Al Viro 已提交
3277
	ctx->pos = PID_MAX_LIMIT + TGID_OFFSET;
3278 3279
	return 0;
}
L
Linus Torvalds 已提交
3280

3281 3282 3283 3284 3285 3286 3287 3288 3289 3290 3291 3292 3293 3294 3295 3296 3297 3298 3299 3300 3301 3302 3303 3304 3305 3306 3307 3308 3309 3310 3311 3312 3313 3314 3315 3316 3317 3318
/*
 * proc_tid_comm_permission is a special permission function exclusively
 * used for the node /proc/<pid>/task/<tid>/comm.
 * It bypasses generic permission checks in the case where a task of the same
 * task group attempts to access the node.
 * The rationale behind this is that glibc and bionic access this node for
 * cross thread naming (pthread_set/getname_np(!self)). However, if
 * PR_SET_DUMPABLE gets set to 0 this node among others becomes uid=0 gid=0,
 * which locks out the cross thread naming implementation.
 * This function makes sure that the node is always accessible for members of
 * same thread group.
 */
static int proc_tid_comm_permission(struct inode *inode, int mask)
{
	bool is_same_tgroup;
	struct task_struct *task;

	task = get_proc_task(inode);
	if (!task)
		return -ESRCH;
	is_same_tgroup = same_thread_group(current, task);
	put_task_struct(task);

	if (likely(is_same_tgroup && !(mask & MAY_EXEC))) {
		/* This file (/proc/<pid>/task/<tid>/comm) can always be
		 * read or written by the members of the corresponding
		 * thread group.
		 */
		return 0;
	}

	return generic_permission(inode, mask);
}

static const struct inode_operations proc_tid_comm_inode_operations = {
		.permission = proc_tid_comm_permission,
};

3319 3320 3321
/*
 * Tasks
 */
3322
static const struct pid_entry tid_base_stuff[] = {
A
Alexey Dobriyan 已提交
3323
	DIR("fd",        S_IRUSR|S_IXUSR, proc_fd_inode_operations, proc_fd_operations),
J
Jerome Marchand 已提交
3324
	DIR("fdinfo",    S_IRUSR|S_IXUSR, proc_fdinfo_inode_operations, proc_fdinfo_operations),
3325
	DIR("ns",	 S_IRUSR|S_IXUGO, proc_ns_dir_inode_operations, proc_ns_dir_operations),
3326 3327 3328
#ifdef CONFIG_NET
	DIR("net",        S_IRUGO|S_IXUGO, proc_net_inode_operations, proc_net_operations),
#endif
A
Alexey Dobriyan 已提交
3329
	REG("environ",   S_IRUSR, proc_environ_operations),
A
Al Viro 已提交
3330
	REG("auxv",      S_IRUSR, proc_auxv_operations),
A
Alexey Dobriyan 已提交
3331
	ONE("status",    S_IRUGO, proc_pid_status),
3332
	ONE("personality", S_IRUSR, proc_pid_personality),
3333
	ONE("limits",	 S_IRUGO, proc_pid_limits),
I
Ingo Molnar 已提交
3334
#ifdef CONFIG_SCHED_DEBUG
A
Alexey Dobriyan 已提交
3335
	REG("sched",     S_IRUGO|S_IWUSR, proc_pid_sched_operations),
R
Roland McGrath 已提交
3336
#endif
3337 3338 3339
	NOD("comm",      S_IFREG|S_IRUGO|S_IWUSR,
			 &proc_tid_comm_inode_operations,
			 &proc_pid_set_comm_operations, {}),
R
Roland McGrath 已提交
3340
#ifdef CONFIG_HAVE_ARCH_TRACEHOOK
3341
	ONE("syscall",   S_IRUSR, proc_pid_syscall),
I
Ingo Molnar 已提交
3342
#endif
3343
	REG("cmdline",   S_IRUGO, proc_pid_cmdline_ops),
A
Alexey Dobriyan 已提交
3344 3345
	ONE("stat",      S_IRUGO, proc_tid_stat),
	ONE("statm",     S_IRUGO, proc_pid_statm),
3346
	REG("maps",      S_IRUGO, proc_pid_maps_operations),
3347
#ifdef CONFIG_PROC_CHILDREN
3348 3349
	REG("children",  S_IRUGO, proc_tid_children_operations),
#endif
3350
#ifdef CONFIG_NUMA
3351
	REG("numa_maps", S_IRUGO, proc_pid_numa_maps_operations),
3352
#endif
A
Alexey Dobriyan 已提交
3353 3354 3355 3356 3357 3358
	REG("mem",       S_IRUSR|S_IWUSR, proc_mem_operations),
	LNK("cwd",       proc_cwd_link),
	LNK("root",      proc_root_link),
	LNK("exe",       proc_exe_link),
	REG("mounts",    S_IRUGO, proc_mounts_operations),
	REG("mountinfo",  S_IRUGO, proc_mountinfo_operations),
3359
#ifdef CONFIG_PROC_PAGE_MONITOR
A
Alexey Dobriyan 已提交
3360
	REG("clear_refs", S_IWUSR, proc_clear_refs_operations),
3361
	REG("smaps",     S_IRUGO, proc_pid_smaps_operations),
3362
	REG("smaps_rollup", S_IRUGO, proc_pid_smaps_rollup_operations),
3363
	REG("pagemap",    S_IRUSR, proc_pagemap_operations),
3364 3365
#endif
#ifdef CONFIG_SECURITY
A
Alexey Dobriyan 已提交
3366
	DIR("attr",      S_IRUGO|S_IXUGO, proc_attr_dir_inode_operations, proc_attr_dir_operations),
3367 3368
#endif
#ifdef CONFIG_KALLSYMS
3369
	ONE("wchan",     S_IRUGO, proc_pid_wchan),
3370
#endif
K
Ken Chen 已提交
3371
#ifdef CONFIG_STACKTRACE
3372
	ONE("stack",      S_IRUSR, proc_pid_stack),
3373
#endif
3374
#ifdef CONFIG_SCHED_INFO
3375
	ONE("schedstat", S_IRUGO, proc_pid_schedstat),
3376
#endif
A
Arjan van de Ven 已提交
3377
#ifdef CONFIG_LATENCYTOP
A
Alexey Dobriyan 已提交
3378
	REG("latency",  S_IRUGO, proc_lstats_operations),
A
Arjan van de Ven 已提交
3379
#endif
3380
#ifdef CONFIG_PROC_PID_CPUSET
Z
Zefan Li 已提交
3381
	ONE("cpuset",    S_IRUGO, proc_cpuset_show),
3382 3383
#endif
#ifdef CONFIG_CGROUPS
Z
Zefan Li 已提交
3384
	ONE("cgroup",  S_IRUGO, proc_cgroup_show),
3385
#endif
3386
	ONE("oom_score", S_IRUGO, proc_oom_score),
3387
	REG("oom_adj",   S_IRUGO|S_IWUSR, proc_oom_adj_operations),
D
David Rientjes 已提交
3388
	REG("oom_score_adj", S_IRUGO|S_IWUSR, proc_oom_score_adj_operations),
3389
#ifdef CONFIG_AUDITSYSCALL
A
Alexey Dobriyan 已提交
3390
	REG("loginuid",  S_IWUSR|S_IRUGO, proc_loginuid_operations),
3391
	REG("sessionid",  S_IRUGO, proc_sessionid_operations),
3392
#endif
3393
#ifdef CONFIG_FAULT_INJECTION
A
Alexey Dobriyan 已提交
3394
	REG("make-it-fail", S_IRUGO|S_IWUSR, proc_fault_inject_operations),
3395
	REG("fail-nth", 0644, proc_fail_nth_operations),
3396
#endif
3397
#ifdef CONFIG_TASK_IO_ACCOUNTING
3398
	ONE("io",	S_IRUSR, proc_tid_io_accounting),
3399
#endif
3400 3401 3402
#ifdef CONFIG_USER_NS
	REG("uid_map",    S_IRUGO|S_IWUSR, proc_uid_map_operations),
	REG("gid_map",    S_IRUGO|S_IWUSR, proc_gid_map_operations),
3403
	REG("projid_map", S_IRUGO|S_IWUSR, proc_projid_map_operations),
3404
	REG("setgroups",  S_IRUGO|S_IWUSR, proc_setgroups_operations),
3405
#endif
3406 3407 3408
#ifdef CONFIG_LIVEPATCH
	ONE("patch_state",  S_IRUSR, proc_pid_patch_state),
#endif
3409 3410
};

A
Al Viro 已提交
3411
static int proc_tid_base_readdir(struct file *file, struct dir_context *ctx)
3412
{
A
Al Viro 已提交
3413 3414
	return proc_pident_readdir(file, ctx,
				   tid_base_stuff, ARRAY_SIZE(tid_base_stuff));
3415 3416
}

A
Al Viro 已提交
3417 3418
static struct dentry *proc_tid_base_lookup(struct inode *dir, struct dentry *dentry, unsigned int flags)
{
3419 3420
	return proc_pident_lookup(dir, dentry,
				  tid_base_stuff, ARRAY_SIZE(tid_base_stuff));
3421 3422
}

3423
static const struct file_operations proc_tid_base_operations = {
3424
	.read		= generic_read_dir,
3425 3426
	.iterate_shared	= proc_tid_base_readdir,
	.llseek		= generic_file_llseek,
3427 3428
};

3429
static const struct inode_operations proc_tid_base_inode_operations = {
3430 3431 3432 3433 3434
	.lookup		= proc_tid_base_lookup,
	.getattr	= pid_getattr,
	.setattr	= proc_setattr,
};

3435 3436
static struct dentry *proc_task_instantiate(struct dentry *dentry,
	struct task_struct *task, const void *ptr)
3437 3438
{
	struct inode *inode;
3439
	inode = proc_pid_make_inode(dentry->d_sb, task, S_IFDIR | S_IRUGO | S_IXUGO);
3440
	if (!inode)
3441
		return ERR_PTR(-ENOENT);
3442

3443 3444
	inode->i_op = &proc_tid_base_inode_operations;
	inode->i_fop = &proc_tid_base_operations;
3445
	inode->i_flags |= S_IMMUTABLE;
3446

3447
	set_nlink(inode, nlink_tid);
3448
	pid_update_inode(task, inode);
3449

3450
	d_set_d_op(dentry, &pid_dentry_operations);
3451
	return d_splice_alias(inode, dentry);
3452 3453
}

A
Al Viro 已提交
3454
static struct dentry *proc_task_lookup(struct inode *dir, struct dentry * dentry, unsigned int flags)
3455 3456 3457 3458
{
	struct task_struct *task;
	struct task_struct *leader = get_proc_task(dir);
	unsigned tid;
3459
	struct pid_namespace *ns;
3460
	struct dentry *result = ERR_PTR(-ENOENT);
3461 3462 3463 3464

	if (!leader)
		goto out_no_task;

3465
	tid = name_to_int(&dentry->d_name);
3466 3467 3468
	if (tid == ~0U)
		goto out;

3469
	ns = dentry->d_sb->s_fs_info;
3470
	rcu_read_lock();
3471
	task = find_task_by_pid_ns(tid, ns);
3472 3473 3474 3475 3476
	if (task)
		get_task_struct(task);
	rcu_read_unlock();
	if (!task)
		goto out;
3477
	if (!same_thread_group(leader, task))
3478 3479
		goto out_drop_task;

3480
	result = proc_task_instantiate(dentry, task, NULL);
3481 3482 3483 3484 3485
out_drop_task:
	put_task_struct(task);
out:
	put_task_struct(leader);
out_no_task:
3486
	return result;
3487 3488
}

3489 3490 3491 3492 3493 3494 3495 3496 3497 3498 3499 3500
/*
 * Find the first tid of a thread group to return to user space.
 *
 * Usually this is just the thread group leader, but if the users
 * buffer was too small or there was a seek into the middle of the
 * directory we have more work todo.
 *
 * In the case of a short read we start with find_task_by_pid.
 *
 * In the case of a seek we start with the leader and walk nr
 * threads past it.
 */
3501 3502
static struct task_struct *first_tid(struct pid *pid, int tid, loff_t f_pos,
					struct pid_namespace *ns)
3503
{
3504
	struct task_struct *pos, *task;
3505 3506 3507 3508
	unsigned long nr = f_pos;

	if (nr != f_pos)	/* 32bit overflow? */
		return NULL;
L
Linus Torvalds 已提交
3509

3510
	rcu_read_lock();
3511 3512 3513 3514 3515
	task = pid_task(pid, PIDTYPE_PID);
	if (!task)
		goto fail;

	/* Attempt to start with the tid of a thread */
3516
	if (tid && nr) {
3517
		pos = find_task_by_pid_ns(tid, ns);
3518
		if (pos && same_thread_group(pos, task))
O
Oleg Nesterov 已提交
3519
			goto found;
3520
	}
L
Linus Torvalds 已提交
3521

3522
	/* If nr exceeds the number of threads there is nothing todo */
3523
	if (nr >= get_nr_threads(task))
3524
		goto fail;
L
Linus Torvalds 已提交
3525

O
Oleg Nesterov 已提交
3526 3527
	/* If we haven't found our starting place yet start
	 * with the leader and walk nr threads forward.
3528
	 */
3529
	pos = task = task->group_leader;
3530
	do {
3531
		if (!nr--)
3532
			goto found;
3533
	} while_each_thread(task, pos);
3534 3535 3536
fail:
	pos = NULL;
	goto out;
O
Oleg Nesterov 已提交
3537 3538 3539
found:
	get_task_struct(pos);
out:
3540
	rcu_read_unlock();
3541 3542 3543 3544 3545 3546 3547 3548 3549 3550 3551
	return pos;
}

/*
 * Find the next thread in the thread list.
 * Return NULL if there is an error or no next thread.
 *
 * The reference to the input task_struct is released.
 */
static struct task_struct *next_tid(struct task_struct *start)
{
O
Oleg Nesterov 已提交
3552
	struct task_struct *pos = NULL;
3553
	rcu_read_lock();
O
Oleg Nesterov 已提交
3554
	if (pid_alive(start)) {
3555
		pos = next_thread(start);
O
Oleg Nesterov 已提交
3556 3557 3558 3559 3560
		if (thread_group_leader(pos))
			pos = NULL;
		else
			get_task_struct(pos);
	}
3561
	rcu_read_unlock();
3562 3563
	put_task_struct(start);
	return pos;
L
Linus Torvalds 已提交
3564 3565 3566
}

/* for the /proc/TGID/task/ directories */
A
Al Viro 已提交
3567
static int proc_task_readdir(struct file *file, struct dir_context *ctx)
L
Linus Torvalds 已提交
3568
{
3569 3570
	struct inode *inode = file_inode(file);
	struct task_struct *task;
3571
	struct pid_namespace *ns;
A
Al Viro 已提交
3572
	int tid;
L
Linus Torvalds 已提交
3573

3574
	if (proc_inode_is_dead(inode))
A
Al Viro 已提交
3575
		return -ENOENT;
L
Linus Torvalds 已提交
3576

A
Al Viro 已提交
3577
	if (!dir_emit_dots(file, ctx))
3578
		return 0;
L
Linus Torvalds 已提交
3579

3580 3581 3582
	/* f_version caches the tgid value that the last readdir call couldn't
	 * return. lseek aka telldir automagically resets f_version to 0.
	 */
3583
	ns = proc_pid_ns(inode);
A
Al Viro 已提交
3584 3585
	tid = (int)file->f_version;
	file->f_version = 0;
3586
	for (task = first_tid(proc_pid(inode), tid, ctx->pos - 2, ns);
3587
	     task;
A
Al Viro 已提交
3588
	     task = next_tid(task), ctx->pos++) {
3589
		char name[10 + 1];
3590
		unsigned int len;
3591
		tid = task_pid_nr_ns(task, ns);
3592
		len = snprintf(name, sizeof(name), "%u", tid);
A
Al Viro 已提交
3593 3594
		if (!proc_fill_cache(file, ctx, name, len,
				proc_task_instantiate, task, NULL)) {
3595 3596
			/* returning this tgid failed, save it as the first
			 * pid for the next readir call */
A
Al Viro 已提交
3597
			file->f_version = (u64)tid;
3598
			put_task_struct(task);
L
Linus Torvalds 已提交
3599
			break;
3600
		}
L
Linus Torvalds 已提交
3601
	}
3602

A
Al Viro 已提交
3603
	return 0;
L
Linus Torvalds 已提交
3604
}
3605

3606 3607
static int proc_task_getattr(const struct path *path, struct kstat *stat,
			     u32 request_mask, unsigned int query_flags)
3608
{
3609
	struct inode *inode = d_inode(path->dentry);
3610
	struct task_struct *p = get_proc_task(inode);
3611 3612
	generic_fillattr(inode, stat);

3613 3614 3615
	if (p) {
		stat->nlink += get_nr_threads(p);
		put_task_struct(p);
3616 3617 3618 3619
	}

	return 0;
}
3620

3621
static const struct inode_operations proc_task_inode_operations = {
3622 3623 3624
	.lookup		= proc_task_lookup,
	.getattr	= proc_task_getattr,
	.setattr	= proc_setattr,
3625
	.permission	= proc_pid_permission,
3626 3627
};

3628
static const struct file_operations proc_task_operations = {
3629
	.read		= generic_read_dir,
3630 3631
	.iterate_shared	= proc_task_readdir,
	.llseek		= generic_file_llseek,
3632
};
3633 3634 3635 3636 3637 3638

void __init set_proc_pid_nlink(void)
{
	nlink_tid = pid_entry_nlink(tid_base_stuff, ARRAY_SIZE(tid_base_stuff));
	nlink_tgid = pid_entry_nlink(tgid_base_stuff, ARRAY_SIZE(tgid_base_stuff));
}